Fender Frontman Reverb guitar amp repair

[Grapsus]

Hi there,

Here is my experience of repairing a small guitar amp. I think it's finally good but I still have a few questions about the circuit.

I got the amp in last December for 10 euros, sold as non working. The transformer was dead, the primary was an open circuit. Somebody apparently already had a go at trying to repair the amp since many wires were cut (first picture). I tried powering the device by injecting DC directly into the circuit with a lab power supply and it worked.

So I thought that ordering a new -/+15 VAC transformer would be enough. What I didn't know is that a transformer may deliver a significantly bigger voltage when it's not fully loaded. Moreover the transformer I bought had a higher power rating, and therefore a higher unloaded voltage. I think this voltage difference was just enough to go above the absolute maximum supply rating of the main amplifying IC, a TDA2050 which literally vaporized. At the time I was totally ignorant about transformers, so I though that it was Farnell's fault delivering me a 18 VAC transformer instead of a 15 VAC one  . (You can learn more about my stupidity in a topic I opened here at the time https://www.eevblog.com/forum/chat/hopeless-problem-with-farnell-need-advise/)

When I finally understood what happened, I ordered a replacement TDA2050. In order not to bust the new one,  I placed two 1N4007 diodes in series with each power rail so that around 1.4 Volts would be shaved on each side. With that dirty hack, the amp finally worked ! Now I think that the design of the amp is stupid: according to the schematic, the TDA2050 is fed with unregulated, just rectified -/+23 Volts when the maximum permitted supply is 25 Volts. Even with the original transformer a 10% increase in mains voltage could have fried the IC. The second picture shows the amp with the new transformer. Since it is bigger than the old one, I mounted it on the bottom of the case.

When I started testing the amp, I found a new problem, the reverb setting had absolutely zero effect on the sound. To troubleshoot this new issue, I injected a clean 440 Hz 5 mVrms sine signal as suggested by TP7 on the schematic. Then I followed all the test points by probing them with a scope and everything agreed within 10% with the schematic. The only weird thing was the amplitude on the output of the reverb tank: the schematic said "10 mVac min" and I was getting like 200 mVrms. I spent hours trying to test everything in the circuit, all being normal and I still had no reverb effect. Finally I decided to remove the reverb tank from the case, it was all glued in an ugly way with double sided tape and foam. What I noticed was that the laminations on each side of the tank were touching the paper case. So I removed the metal base from the paper case which was also glued in an awful way. Finally with the "naked" reverb tank I had reverb in the output sound ! The amplitude at the output of the tank went down to 20 mVrms which seems a lot closer to "10 mVac min". So I guess the problem was really mechanical and not electronic, by touching the case, the tank must have been propagating the wave in a different mode. I carefully reassembled everything and it kept working 

In the last picture you'll find an oscilloscope screen shot with the sound starting to come in the tank and getting out around 25 ms later. If anyone has experience with reverb tanks, can you tell me if it looks as it should ?

Another question I have is about the volume. On the schematic you can see that the input can follow two different paths, clean and overdrive. After the channel switch S1, the signal is buffered by U3B and its output should be 410 mVrms. Obviously this figure will depend on the position on volume and gain pots. With the overdrive channel, if I set the gain on 1 (no visible distortion) and the volume about in the middle, TP13 is indeed around 400 mVrms. Now if I switch to the clean channel and I vary the volume knob,
TP13 only goes from 0 to 100 mVrms when the volume is at the max. Do you think it is normal for the clean channel to be less loud than the overdrive or is something defective in the circuit ? The clean channel is only composed of one op amp U1A and some passive components. I don't understand the circuit around U1A, but I tried to simulate it with LTspice and I get at most 80 mVrms on the output wich agrees with my measurements.

Schematic: http://support.fender.com/schematics/guitar_amplifiers/Frontman_15G_15R_schematic.pdf
Different amp but the same reverb circuit which is absent from the first schematic: http://support.fender.com/schematics/guitar_amplifiers/Bullet_Reverb_schematic.pdf

[taxman]

Graspus

I think you did a great job. I know a lot less about repair than you, but I know guitars.  It is most common for an overdrive channel to be louder than the clean channel.  That is really what overdrive means.  This is frequently done by adding an additional amplifier stage to the overdrive channel.

[Chipmunk]

Great job for linking to the original thread too, the lesson you learned YOU won't forget for sure, and maybe it will save someone else

The same thing can happen with DC power bricks (the old transformer type, switchmode tend to be a lot better). If you use one that's got a much higher current rating, the unregulated type can have a horrifyingly huge DC output. I've seen 9vDC ones produce well over 13 volts

[Grapsus]

Graspus

I think you did a great job. I know a lot less about repair than you, but I know guitars.  It is most common for an overdrive channel to be louder than the clean channel.  That is really what overdrive means.  This is frequently done by adding an additional amplifier stage to the overdrive channel.

Ok, I understand: the clean channel keeps a low gain so it doesn't allow harmonic distortion and overdrive is more like "let's allow the user to saturate every possible component in the amp"  In the circuit diagram, it would have been better if they specified ranges for each channel on the test points instead of just printing one figure.

I'm a total beginner with music, but this amp has a much more pleasant sound than the tiny Epiphone amp that came with my guitar so I think I'm going to keep it.
Plus the build quality is pretty good except for powering the TDA2050 with unregulated voltages. Looks like they only use prime spec parts, Nippon Chemicon caps, TI and Rohm op amps etc.

Great job for linking to the original thread too, the lesson you learned YOU won't forget for sure, and maybe it will save someone else

The same thing can happen with DC power bricks (the old transformer type, switchmode tend to be a lot better). If you use one that's got a much higher current rating, the unregulated type can have a horrifyingly huge DC output. I've seen 9vDC ones produce well over 13 volts

Yeah I still feel bad thinking about it  But, even if it took time, I learned a lot repairing this amp, about transformers, reverb delay lines and audio signal in general.

[commongrounder]

Just a comment on using unregulated voltage on the TDA power amp chip.  These chips use a classic complimentary output transistor design which has very high inherent DC ripple rejection.  One wants to design an amp circuit so it is connected as closely as possible to the "power reservoir", the filter capacitors, so the transient response and output power can be as high as possible.  Regulating the voltage source can be done, but the regulator circuit has to be very sophisticated to prevent instability in the power amp section.  In a nutshell, the Fender circuit isn't really a "shortcut" or cost saving measure, but them following the data sheet recommendations for the chip.
Congratulations on getting your amp up and running!

[Grapsus]

Just a comment on using unregulated voltage on the TDA power amp chip.  These chips use a classic complimentary output transistor design which has very high inherent DC ripple rejection.  One wants to design an amp circuit so it is connected as closely as possible to the "power reservoir", the filter capacitors, so the transient response and output power can be as high as possible.  Regulating the voltage source can be done, but the regulator circuit has to be very sophisticated to prevent instability in the power amp section.  In a nutshell, the Fender circuit isn't really a "shortcut" or cost saving measure, but them following the data sheet recommendations for the chip.
Congratulations on getting your amp up and running!

I have nothing against using unregulated power for the TDA chip, it's just that this unregulated voltage seems dangerously close to the absolute maximum. Any variation on the AC line and it can go out in fire. On the other hand, using a lower voltage transformer would imply less power when the volume is maxed, so they tried to get the maximum out of this chip.