Home Projects Car About

Carver TFM-25 Amplifier

A story of how three broken Carver TFM-25 amplifiers became two working ones.

The TFM-25 with the lid removed. The supply capacitors are much larger than that of the TFM-15, and each channel sports ten output transistors (as opposed to four). The amp is capable of delivering 225W x 2 at 8ohms, 350W x 2 at 4ohms, and 350W bridged into a single 8ohm load.

A photo from the back. Note the switch that toggles channel bridging.

The rear panel removed.

The PCB and transistor heatsinks separate completely from the amp’s chassis. A large AC transformer and the meter panel are all that remain.

The transformer and chassis.

A closer look at the amp’s main PCB. This component is from the $10 TFM-25 I acquired in 2005 from “Suthin’s Electronics”, a local repair shop in Bethlehem, PA. Close inspection of the chassis and PCB showed a severe dent, as well as a snapped circuit board, in the bottom-left corner of the amp. The amp had probably been dropped by the previous owner, and repairing the PCB should fix the problem. Unfortunately, the amp had been mined for parts (the output transistors, in particular), so the restoration project was put on hold until some suitable replacements could be found.

My luck changed over winter break that year after locating a second broken TFM-25 at “Electronic Shop” near Purdue University in West Lafayette, IN. I purchased it and a broken TFM-15 amp (successfully repaired here) for $50. This TFM-25, however, was in horrible shape. It suffered from several exploded power diodes, disintegrated power resistors, melted wiring, and a burnt-up transformer, most likely due to a lightning strike. However, the parts that weren’t destroyed would make great replacements for the other amp! The chassis, being dent free, was kept along with other salvaged parts (placed inside the bag in the photo).

Back to the snapped PCB of the first amp. All traces appearing cracked on the board were bridged, and all the on-board components (aside from the capacitors) were checked using a standard multimeter.

A close-up of the broken corner, with the components bridged using breadboard wire.

The traces around one of the supply caps were completely burnt-up as well. The photo shows how the components were bridged.

Given what seemed like an easy fix, it turned out not to be the case. Despite repeated testing, the amp would not turn on without a loud buzz from the transformer and smoke appearing from the right channel’s power diodes. Since bridging the PCB seemed to be a failure, I decided to move a full set of working parts over to the other amp’s PCB board.

The one working transformer from the dropped amp is installed in the chassis. A complete set of output transformers is screwed in as well.

More than two dozen broken parts were replaced, ranging from power diodes, to resistors, to transistors. Several of the white power resistors in the photo had to be replaced since they were completely burnt up. Also, one of the thermal trip sensors needed to be swapped for a working one. Despite extensive testing and retesting, the amp showed exactly the same problems as before! The transistor would hum loudly, and smoke would emanate from the right channel power diodes.

Despite starting this project nearly two years ago, I decided to revisit it after ordering another broken TFM-25 amp on eBay. In addition to ordering a TFM-25 service manual, I arranged to have Lehigh University’s electronics lab equipment at my disposal.

Using a VariAC device (which linearly changes the voltage applied to a device from an AC socket), it became apparent that one of the channels would become incredibly hot after a minute or so around 80VAC. At this voltage, the transistor would start buzzing, too.

Because this problem affected only one of the channels, the output transistors were again reinvestigated. They were swapped around to try and resolve the issue, but it brought no success.

Back to the drawing board! Capacitors which could not be tested before were de-soldered and tested using lab equipment. Each on-board transistor was removed and re-checked. Careful PCB examination around one of the transistors in the TRIAC region (the circuitry controlling the power supply) showed a hairline crack along a PCB trace. Could this be the problem the root of the problem?

Success! The meters light up to a two-click salvo from the output relays! The amp is hooked up to a sinusoidal signal generator to see what its frequency response turns out to be after this repair nightmare.

Much to my surprise, the output is clean and stable! A pair of dummy 10ohm power resistors were used as a load on each channel, and both were driven and scoped simultaneously.

Frequency response and gain at 1kHz look fantastic!

The amp maintains signal gain even for frequencies below 20Hz. My work with this amp is done!

After this successful repair, my attention shifted to the amp I ordered from eBay. It was being sold broken and “as-is” by the seller. I won the bidding for $129.01 shipped.

To my surprise, the seller shipped the item with $40 in cash attached to a note. Apparently, his shipping company had scraped the bottom of the amp and bent in one of the rack handles. The offer was more than adequate compensation! A few hours in a vice had the rack handles and faceplate looking like new again.

Inside the amp. The PCB was not nearly as dusty as the other one, and a visual inspection showed no damaged parts. What luck!

A thorough examination of the PCB showed only a single component at fault: one of the small blue and red diodes in the center of the photograph. I replace the pair for good measure.

This second amp was manufactured a couple of years prior to the other TFM-25. The sub-standard speaker posts were replaced with a spare pair that matched the other amp.

Another success! The amp turns on and operates wonderfully without nearly any of the sweat and blood from the other repair.

One issue that cropped up was the seemingly random deactivation of the output relays after several minutes of play. Usually a slight bang on the chassis would cause them to switch back on. I replaced the relays with the spare ones on-hand to determine if they were the problem.

The TFM-25 service manual mentioned the bias calibration procedure, and this was performed on both amps. The relay clicks disappeared! I also calibrated both meters, so the amps now run identically in-spec.

The two TFM-25 amps, closed up and ready for installation.

The TFM-25 far exceed the power output of my TFM-15 and TFM-15CB. This makes them much more capable of driving inefficient speakers. The extra headroom and power output is also great when I use these amps to deejay!


Carver TFM-25 (Suthin's Electronics) $10.00
Carver TFM-25 (Electronic Shop) $25.00
Carver TFM-25 (Ebay) $89.01
Service manual $12.00
Solder and wick $9.00

Total $145.01