The issue presented is whether Computer Displays International, Inc. (“CDI”) violated a consent decree entered into between CDI and Motorola, Inc. on June 24, 1982. The district court found a violation and we affirm.
I.
The plaintiff, Motorola, manufactures a variety of electronic products, including cathode ray tube (“CRT”) display monitors.1 One of Motorola’s display monitor product lines is the Model DS 110° monitor (“DS”). Motorola first displayed the DS monitor in a private hotel suite during the June 1980 National Computer Conference in Anaheim, California. Single engineering samples of the DS monitor were thereafter sold or given to 50-75 selected Motorola customers. On January 19, 1981, Motorola obtained a copyright for the DS monitor service manual, which contained a service schematic diagram. The DS monitor, however, was not introduced to the public until the May 1981 National Computer Conference in Chicago.
While Motorola was developing the DS series line, the three individual defendants were Motorola employees. Robert Gatza was the Director of Marketing for Motorola’s Display Operations; Chris Petri worked directly under Gatza at Motorola; and Thomas Fair was the Purchasing Agent for Motorola’s Electrical Purchasing Group. On October 13, 1980, Gatza’s superior informed him that he should seek other employment because of a personality conflict; Gatza was retained on the Motorola payroll, however, through February 28, 1981. Rather than join another company, Gatza and Petri incorporated CDI in November, 1980. When Motorola learned of Petri’s involvement with Gatza and CDI, Petri was also terminated effective February 28, 1981. Fair resigned from Motorola on June 29, 1981 to take a position with CDI.
*1152In January, 1981, Gatza, using a pseudonym, gave the parameters for a display monitor deflection yoke2 to a Motorola supplier, DOK of America, Inc., for a quotation on the cost of producing such a yoke. On February 23, Gatza approached another Motorola supplier, Electronic Components Corp., concerning the manufacture and supply of a power transformer.3 The specifications given by Gatza for both the yoke and the power transformer were identical to those used for the same components in Motorola’s DS monitor. CDI surreptitiously obtained a complete DS monitor in early February and began reverse engineering4 from the DS monitor in early March. In May, CDI began marketing and selling its Model MPG 110° display monitor (“MPG”), which was substantially identical to Motorola’s DS monitor.
As a result of the great similarity between the DS and MPG monitors, Motorola filed súit against CDI alleging trademark and copyright infringement, breach of employment contracts, and misappropriation of confidential information. This action terminated with the signing of a consent decree on June 24, 1982. The consent decree provided, in relevant part:
7. Robert G. Gatza, Chris A. Petri and Thomas S. Fair ..., by reason of their prior employment with Motorola Display Systems Unit in positions of trust and confidence collectively had access to and knowledge of that Motorola confidential and/or proprietary information respecting research, design, development, manufacturing, sources of supply, marketing and business plans for cathode ray tube display modules in general, and Motorola Model DS3000 and DS4000 series display product lines in particular (hereinafter “said Motorola information”).
9. Defendants ... are permanently enjoined from: (a) using, publishing or .otherwise disclosing to others said Motorola information; .'.. and (e) engaging in business activities with respect to cathode ray tube display modules that will inevitably result in, or pose a substantial risk of the use or disclosure of said Motorola information.
10. Defendants are permanently enjoined from manufacturing, using and selling the display modules currently designated by CDI as its Model MPG series and all models embodying Motorola information ____
The decree did not bar CDI from introducing a new 110° display monitor so long as it was not the MPG series monitor and did not embody “Motorola information.” Indeed, a settlement agreement signed the same day as the consent decree specifically allowed CDI to introduce a new 110° monitor after October 1, 1982. To facilitate protection of Motorola confidential and/or proprietary information, CDI agreed to submit a prototype of any new model thirty days in advance of its introduction. Subsequently, CDI informed Motorola that it would begin marketing a new monitor (the Model CDI monitor) on October 1, 1982.
CDI shipped a prototype of the Model CDI to Motorola on July 30, but Motorola refused to approve the new Model CDI. In a letter dated August 27, Motorola informed CDI that it found “the circuitry utilized in the next generation monitor [the Model CDI] ... [to be] essentially the same as the prior MPG series monitors.” Negotiations between Motorola and CDI ensued, but no agreement was reached by the end of September. When it became apparent that CDI would market the Model CDI *1153without Motorola’s approval, Motorola filed a motion with the district court on September 28, asking for a temporary restraining order on the sale of the Model CDI monitor and for a rule to show cause why CDI should not be found in contempt of the June 24 consent decree.
On October 1, the district court held a hearing on Motorola’s motion, at which Motorola and CDI presented testimony on the identity of the two display monitors. The district court denied Motorola’s motion for a temporary restraining order. The court also ruled preliminarily that Motorola had not borne its burden of proving that the monitors were the same, but it continued the hearing on this issue to October 25. The district court did warn CDI:
I think the defendant CDI is at risk. In other words, I have not found that you are not’ violating the injunction or that the products are different. I am saying for the moment on this one-day hearing I have not been persuaded that the plaintiff has sustained its burden of demonstrating ... that the products are the same. There are risks inherent in what you are doing.
Accepting these risks, CDI commenced marketing and selling the Model CDI monitor after the hearing.
Hearings on Motorola’s contempt motion were then held on ten days over a period of three months, during which time the district court heard an additional 1200 pages of testimony and considered numerous exhibits and depositions on the question whether CDI’s new display monitor was barred by the June 24 consent decree. The district court ruled from the bench on July 1, 1988 that CDI was in contempt of the June 24 consent decree. The district court asserted two grounds for reaching this conclusion. First, the court applied the patent law doctrine of equivalents and held that the Model CDI monitor was the same as the Model MPG monitor. Second, the court interpreted the consent decree to mean that any information that was confidential and/or proprietary to Motorola at the time the individual defendants left Motorola could not be incorporated into any CDI display monitor even though that information might later become publicly available. Finding that Motorola’s entire DS monitor was confidential when the defendants left Motorola, the court held that the use of substantially similar circuitry in Motorola's DS monitor and CDI’s MPG and Model CDI monitors constituted a use of Motorola confidential information in violation of the consent decree. In addition, the district court found that the deflection yoke, power transformer, and mounting brackets used in both the MPG and Model CDI monitors originated with Motorola’s DS monitor and, thus, also constituted a use of confidential Motorola information.5 Although the district court found CDI in violation of the consent decree, the court expressly reserved the question of compensatory damages due Motorola as a result of that breach. CDI appeals.6
II.
During oral argument, we sua sponte raised the issue of appellate jurisdiction and ordered the parties to file supplemental *1154briefs limited to this issue within ten days.7 Both sides argue that we have appellate jurisdiction, though each asserts a different ground. Motorola contends that our jurisdiction can be based only on the fact that the district court’s contempt order was a final order under 28 U.S.C. § 1291. CDI argues in the alternative that the district court’s order was a final order under § 1291, a collateral order under Cohen v. Beneficial Industrial Loan Corp., 337 U.S. 541, 69 S.Ct. 1221, 93 L.Ed. 1528 (1949), or an order “granting, continuing, [or] modifying ... [an] injunction! ]” under 28 U.S.C. § 1292(a)(1).
A.
We turn first to CDI and Motorola’s argument that the district court’s order was a final order. While it is true that most post-judgment orders are final decisions within the ambit of § 1291, not all are. To be final, the post-judgment order must still dispose completely of the issues raised. Sportmart, Inc. v. Wolverine World Wide, Inc., 601 F.2d 313, 316 (7th Cir.1979). An order finding a party in civil contempt disposes of all the issues raised only if it includes both a finding of contempt and the imposition of a sanction. Shuffler v. Heritage Bank, 720 F.2d 1141, 1145 (9th Cir.1983); United States Steel Corp. v. Fraternal Ass’n of Steel Haulers, 601 F.2d 1269, 1273 (3d Cir.1979); Steinert v. United States, 571 F.2d 1105, 1107 (9th Cir.1978); Major v. Orthopedic Equipment Co., 561 F.2d 1112, 1115 (4th Cir. 1977); SEC v. Naftalin, 460 F.2d 471, 475 (8th Cir.1972). Here, the district court’s order included only a finding of contempt; the district court expressly reserved the question of compensatory damages for a later date. Thus, the finding of contempt was an interlocutory order and nonappealable under § 1291.
B.
CDI contends that even if the district court’s contempt order .was not technically final, it falls within the very narrowly defined collateral order exception to § 1291. To be appealable as a collateral order, the contempt order must meet three conditions: first, it “must conclusively determine the disputed question”; second, it must “resolve an important issue completely separate from the merits of the action”; and third, it must “be effectively unreviewable on appeal from a final judgment.” Coopers & Lybrand v. Livesay, 437 U.S. 463, 468, 98 S.Ct. 2454, 2457, 57 L.Ed.2d 351 (1978). See also Flanagan v. United States, — U.S.-, 104 S.Ct. 1051, 1055, 79 L.Ed.2d 288 (1984); Cohen v. Beneficial Industrial Loan Corp., 337 U.S. 541, 545-47, 69 S.Ct. 1221, 1225-26, 93 L.Ed. 1528 (1949); Rohrer, Hibler & Replogle, Inc. v. Perkins, 728 F.2d 860, 862 (7th Cir.1984) (per curiam). The failure to meet a single condition will bar appellate review under this doctrine.
In the instant case, the district court’s contempt order fails to satisfy two of the three conditions. Whether CDI was in contempt of the June 24 consent decree was the principal issue before the district court. Thus, the contempt order did not resolve an “issue completely separate from the merits of the action.” In addition, the district court’s finding of contempt would be reviewable on appeal from a final judgment —i.e., an order that found CDI in contempt and imposed a sanction for that contempt. We hold, therefore, that the contempt order is not appealable as a collateral order.
C.
CDI’s final argument is, that the contempt order is reviewable as an interlocutory order “granting, continuing, [or] modifying ... [an] injunction[ ] ____” 28 U.S.C. § 1292(a)(1). We find no merit to CDI’s contention that the contempt order continued the original injunction in the June 24 consent decree or that it was a temporary injunction pending resolution of *1155the damages question. We limit our discussion, therefore, to whether the contempt order modified or interpreted the earlier injunction.
CDI argues that, while the district court may not have changed the language of the consent decree injunction, the court’s interpretation of that injunction was so expansive as to be, in effect, a modification of it. The distinction between an order interpreting an injunction and one modifying an injunction is not always clear, but the distinction determines our jurisdiction.8 Orders that merely interpret or clarify an injunction are not appealable under § 1292(a)(1). See Buckhanon v. Percy, 708 F.2d 1209, 1212 (7th Cir.1983); Sperry Corp. v. Minneapolis, 680 F.2d 1234, 1236 (8th Cir.1982); Major v. Orthopedic Equipment Co., 561 F.2d 1112, 1115 (4th Cir. 1977); SEC v. Investment Corp. of America, 369 F.2d 383, 384 (7th Cir.1966). Orders that modify an injunction, however, are explicitly within the language of § 1292(a)(1) and appealable.
The mere allegation by an appellant that the district court’s order has modified rather than interpreted an injunction will not suffice to vest this court with jurisdiction. But see In re Arthur Treacher’s Franchisee Litigation, 689 F.2d 1150, 1155 n. 9, 1158 n. 14 (3d Cir.1982). To determine our jurisdiction we must look beyond the characterization given the contempt order by the parties and the district court to the actual effect of that order. See Movie Systems, Inc. v. MAD Minneapolis Audio Distributors, 717 F.2d 427, 429 (8th Cir.1983); Bank of California v. Arthur Andersen & Co., 709 F.2d 1174, 1176 (7th Cir.1983); Buckhanon v. Percy, 708 F.2d at 1212; Sperry Corp. v. Minneapolis, 680 F.2d at 1236-37; Major v. Orthopedic Equipment Co., 561 F.2d at 1115; SEC v. Investment Corp. of America, 369 F.2d at 384. Cf. Carson v. American Brands, Inc., 450 U.S. 79, 83, 101 S.Ct. 993, 996, 67 L.Ed.2d 59 (1981) (order having effect of refusing an injunction held appealable under § 1292(a)(1)). The rationale for this approach is clear: an order labeled as an interpretation may substantially alter the legal relationship originally contemplated by the parties and incorporated by them in the consent decree. An erroneous interpretation can have the same effect, and be just as devastating to a party, as an order explicitly modifying or refusing to modify an injunction. A true interpretation, on the other hand, does not change the parties’ original relationship, but merely restates that relationship in new terms. When the district court alters the legal relationship between the parties, whether by “modification” or “interpretation,” we have jurisdiction to review that change.
In some instances, it may be apparent from a cursory examination of the “modifying” order that the appellant’s allegation of a modification is frivolous; in such a case, we will summarily dismiss for lack of appellate jurisdiction. Cf. Buckhanon v. Percy, 708 F.2d 1209 (7th Cir.1983) (cursory review of second order sufficient to find appellate jurisdiction). We will also dismiss if the alleged modification is merely an attempt to relitigate the issuance of the original injunction. In other cases, however, the true effect of the modifying order will not be obvious, thus necessitating a careful review of the merits to determine the order’s effect. This appeal falls into the latter category of cases.9
*1156III.
The first prong of the, consent decree injunction permanently enjoins GDI “from manufacturing, using and selling the display modules currently designated by CDI as its Model MPG series____” Applying the patent law doctrine of equivalents, the district court held that the Model CDI monitor was the “same” as the MPG monitor and, therefore, in violation of the injunction.
With respect to this prong of the consent decree injunction, CDI asserts two errors on appeal. First, CDI argues that the district court modified the injunction to read, in effect, as follows: “Defendants are permanently enjoined from manufacturing, using and selling display modules of the type designated by CDI as its Model MPG series or equivalents thereof as determined by a percentage of similarity test.” 10 Second, CDI argues that the district court clearly erred in finding that the MPG and the Model CDI monitors are the same.
A.
The consent decree states that CDI may no longer sell its Model MPG series monitor. The first question before the district court, then, was what the parties intended to cover by this clause. The district court interpreted the clause to mean that CDI could not sell the MPG monitor or any monitor that was the “same” as the MPG. CDI does not contest this eminently logical interpretation of the decree. Rather, CDI challenges the degree of identity barred and the methodology used by the district court to determine that degree of identity.
The district court did not explicitly define what degree of sameness was barred by the consent decree. The court's references during its oral findings to CDI’s failure to make “substantial changes in the product,” though, suggest that the district court believed the Model CDI was barred if it was substantially like the MPG. To say that the consent decree prohibited only the sale of a display monitor series identical to the MPG in all respects or with only minimal changes would make the injunction meaningless. Even Gatza recognized that more than a trivial change would be required to comply with the decree. Gatza, Tr. 721. Moreover, this interpretation is in accord with trade secret law, which protects the holder of a trade secret from uses substantially derived from that secret. See Forest Laboratories, Inc. v. Pillsbury Co., 452 F.2d 621, 625 (7th Cir.1971) (applying Wisconsin law, but cited with approval in Affiliated Hospital Products, Inc. v. Baldwin, 57 Ill.App.3d 800, 15 Ill.Dec. 528, 373 N.E.2d 1000 (1978)); Filter Dynamics International, Inc. v. Astron Battery, Inc., 19 Ill.App.3d 299, 317, 311 N.E.2d 386, 400 (1974) (plaintiff must show variations are insignificant to recover); Vendo Co. v. Stoner, 105 Ill.App.2d 261, 279, 245 N.E.2d 263, 272 (1969) (vast differences will preclude recovery); Restatement • of Torts § 757, comment c. We find no error in the district court’s interpretation of the scope of the injunction.
The next question for the district court was whether the Model CDI monitor was substantially like the MPG. The consent decree does not specify how this determination is to be made. CDI argues that the two monitors are not substantially alike under the terms of the consent decree if a model series change was made. This argument is specious. Gatza testified that CDI could change the model series, without making any change in the monitor, just to spur sales in a slow market. Gatza, Tr. 757. See also Slavik, Tr. 373. Under CDI’s theory, any change would require a model series change. The real question, however, is not change, but the degree of *1157change. The model series test advanced by CDI does not further this determination. The district court properly rejected CDI’s approach in favor of a doctrine of equivalents analysis. Although CDI contests this use of a patent law doctrine in a trade secret case, we believe that the district court’s approach was appropriate.
In patent law, courts are often called upon to determine whether two devices are substantially alike. They do this through application of the doctrine of equivalents, which “is based on substantial identity between the teachings of the claims of the patent in suit and the accused device in structure, mode of operation and result to be accomplished____” Rothstein v. Atlanta Paper Co., 321 F.2d 90, 95 (5th Cir.1963). If an accused device “performs substantially the same function in substantially the same way to obtain the same result” as the patent claim, the patent is infringed. Sanitary Refrigerator Co. v. Winters, 280 U.S. 30, 42, 50 S.Ct. 9, 13, 74 L.Ed. 147 (1929). See also Graver Tank & Manufacturing Co. v. Linde Air Products, 339 U.S. 605, 608, 70 S.Ct. 854, 856, 94 L.Ed. 1097 (1950); Wolens v. F.W. Woolworth Co., 703 F.2d 983, 990 (7th Cir.1983). While the doctrine of equivalents is strictly a patent law doctrine, its usefulness in determining analogous trade secret cases had been noted and approved. See Cataphote Corp. v. Hudson, 442 F.2d 1290, 1294-95 (5th Cir.1970); Bolt Assocs. v. Alpine Geophysical Assocs., 365 F.2d 742, 748 (3d Cir.1966). Cf. Restatement of Torts § 757, comment c. Use of the doctrine in a trade secret case is particularly appropriate where, as here, the parties have contracted through the consent decree to prohibit the manufacture and sale of a display monitor substantially like the “secret” monitor. Cf Bolt Assocs. v. Alpine Geophysical Assocs., supra, (doctrine of equivalents analogy useful where test is mechanical equivalence). Thus, we hold that the district court properly applied the analogous patent law doctrine of equivalents to effect the parties’ intent as embodied in the consent decree.
B.
Whether the Model CDI monitor is substantially like the MPG monitor is a question of fact. The district court’s determination that the two monitors are substantially alike may be overturned, therefore, only if that finding is clearly erroneous. See Fed.R.Civ.P. 52(a).
In designing the Model CDI monitor, Motorola concedes that CDI made some changes from the MPG. CDI contends that its alterations in the mechanical components (i.e., the chassis pan shape, heat sink locations, and method of attaching the printed circuit board (“PCB”) to the chassis pan), the electromechanical components (i.e., the layout of the PCB), and the thermal components (i.e., heat sinks), make the district court’s finding of substantial identity clearly erroneous. We disagree. The chassis pan shape, location of heat sinks, and method of attaching the PCB relate only to the way the monitor fits together. As Motorola’s expert witness John Renniek testified, “these things have to fit together reasonably well or they wouldn’t operate. The specification outlines what kind of container this whole thing must go into; and if it goes in there, so as to satisfy the specification, exactly how it is held jn place has no real bearing on the differences.” Renniek, Tr. 320. It is equally clear that the different PCB layout was inconsequential. Motorola presented compelling evidence that a circuit, schematic could be given to five layout designers who, applying the rules of design, could produce five different PCB layouts that would work.11 See, e.g., Padgett, Tr. 171; Dols, Tr. 267; Renniek, Tr. 317. Thus, these differences conceded by Motorola neither weaken nor contradict the district court’s finding of substantial identity-
Rather than examine the changes urged by CDI, the district court focused on the lack of change in the circuitry of the two *1158display monitors. Defense witnesses Gatza, Dan Bhagat, LeRoy Carson, and Vincent Formanek all agreed that the heart of any electronic device is its circuits. See Gatza, Tr. 773 (circuits and PCB are heart); Bhagat, Tr. 960 (circuits are important factor); Carson, Tr. 1228 (heart); Formanek, Tr. 1273-74 (circuitry important). CDI insists, though, that the district court’s finding was clearly erroneous even on this basis. ,
Motorola’s expert witnesses noted eight differences (hereinafter “Block 1” through “Block 8”) between the circuit schematics for the MPG and Model CDI. Block 4 was later discovered to be identical, the difference haying resulted from a drawing error.12 Even given these differences, Motorola’s experts still found the Model CDI to be substantially like the MPG.13
Taking the seven actual differences into account, Terry Geisz testified that 60-70% of the circuits in the MPG and Model CDI were exactly alike. Geisz, Tr. 114, 1366 (clarified on rebuttal at 75% exact identity). Where the circuits were not identical, Geisz found that they were equivalent “in terms of means, function and result____” Geisz, Tr. 114. This testimony alone provides the support necessary for us to affirm -the district court’s findings. In addition, John Dols, on cross-examination by Motorola, agreed with Geisz’s estimate of 60-70% exact identity of circuits. Dols, Tr. 243-44. By counting and comparing the number of active devices in each monitor, Rennick concluded that there was a 74% identity between the two monitors’ circuits. Rennick, Tr. 314. When Rennick excluded diodes from his count (some experts consider them trivial components), the degree of identity rose to 82%. Id. Dr. Kenneth Haag compared the circuitry parts lists for each monitor and found that 80% of the items on the MPG list were also on the Model CDI list. Haag, Tr. 455. Based on this testimony, the district court correctly found the MPG and the Model CDI monitors to be substantially alike.
Relying on an off-the-cuff estimate made by Rennick in response to a question from the court, CDI argues that this high degree of identity is inconsequential in light of a 50% identity between all display monitors. See Rennick, Tr. 346. After further study, however, both Rennick and Geisz testified on rebuttal that the similarity between the MPG and other companies’ competitive monitors ranged from 11% (Audiotronics Model 951-01/02) to 35% (Sylvania Model DDM1524). Rennick, Tr. 1344-45; Geisz, Tr. 1366. The MPG and Model CDI monitors were extraordinarily similar.
Further support for the district court’s finding of substantial likeness is found in the testimony concerning the ease of redesigning the circuits in Blocks 1 through 8, CDI’s reverse engineering process, and CDI’s representations to its customers concerning the changes. Rennick, Dr. Steven Burns, and Dr. Haag each testified that the circuitry in Blocks 1 through 8 was the éasiest to redesign; See Rennick, Tr. 316; Burns, Tr. 413; Haag, Tr. 455. Dr. Haag further testified that the engineering books for the Model CDI redesign effort demonstrated that the design work was completed in reverse order. Rather than following the customary design effort from output back to input, the Model CDI design effort went from input to output. Haag, Tr. 462. The notebooks also showed that attempts to alter the Model CDI circuits failed so the designer reverted to circuits used in the MPG. Id., Tr. 465-66, 473, 483-84. Finally, CDI represented to its customers that *1159“[t]he model numbering system has been simplified by dropping the MPG prefix; ... mechanical and electrical dimensions and connections are identical in form, fit, and function to the MPG series____” These additional facts suggest that CDI never intended to undertake a full redesign effort to avoid the proscriptions of the consent decree.
Based on our extensive review of the record, we hold that the district court’s finding that the Model CDI monitor was substantially like the MPG monitor was not clearly erroneous. Indeed, that finding was clearly correct.
IV.
Because we agree with the district court’s determination that the Model CDI was substantially like the MPG, we need not reach the issue whether the Model CDI embodied confidential and/or proprietary Motorola information. While we do not pass judgment on the district court’s resolution of this issue, we must discuss it briefly in order to decide CDI’s “Motion And Memorandum In The Nature Of A Rule 60(b) Motion Requesting An Order That The District Court Grant A Trial On the Single Issue Of Whether The MDS (DS) Monitor Was In The Public Domain By February 28, 1981.”
The confidentiality issue was not raised in Motorola’s motion for a rule to show cause — that motion was limited to the question whether the Model CDI monitor was in essence the MPG monitor. At the continuation of the October 1 hearing on Motorola’s motion, CDI explicitly objected to any broadening of the scope of the contempt proceedings. Tr. 232. The district court noted the objection, but did not rule on it. Instead, the court stated, “When we get to that evidence, let’s see what it is and talk about it again.” Tr. 233. The court also requested that Motorola’s counsel “highlight” the introduction of this evidence so that a ruling could then be made. Tr. 234. Motorola’s counsel did not “highlight” the evidence and no motion was ever made to broaden the scope of the proceedings. Given CDI’s view of the scope of the hearing, it is entirely possible that it did not present, or even attempt to obtain, all of the relevant evidence that it might have wished to present with respect to this issue. We are not in a position to make this determination. We leave it to the district court, therefore, to decide in the first instance whether more evidence on this issue is appropriate. Accordingly; we deny CDI’s motion without prejudice to its renewal in the district court.
V.
For the reasons expressed above, the district court’s finding of contempt is affirmed.