The Repair Bench: Marshall Valvestate 8200 Amplifier Head

| May 30, 2010 | 13 Comments


Learning how to troubleshoot electronic circuits is one of the handiest skills anyone can have if they are building or modifying amplifiers or effect pedals. At some point you’ll power-up a circuit and find that it doesn’t work, and then what? For this reason we write “The Repair Bench” section of Guitar Kit Builder about our own troubleshooting of amplifiers and other devices, to pass along to the reader the thought process, tips and techniques of troubleshooting electronic equipment.

In this episode of The Repair Bench we are troubleshooting a Marshall Valvestate 8200 Amplifier head that we purchased used. The previous owner (PO) offered us the amp, but asked us to name the price. We’re not experts here in valuing amplifiers, but we found three reports on for Valvestate 8200 sales in years past for $225.00 (2010), $192.00 (2007) and $119.00 (2005). We really wanted this amp since we hadn’t previously opened up a Valvestate, and figured it was worth at least $200.00 working. We offered $75.00 and the PO countered at $100 with a Marshall footswitch and two amp cables, and we settled on $90.00. The PO said he had stopped playing guitar some time ago, and that to his recollection the amp had just stopped working at one point and he had never tried to have it repaired.


First, a little about this amp. The 8200 is from Marshall’s Valvestate line, which was introduced in the 1990s as a “revolutionary concept in amplification.” Marshall’s design intent was to emulate the tone of a cranked-to-the-max all-tube power stage with the cost-cutting advantages of solid state technology. The 8200 is the 100×100 watt, bi-chorus model of the Valvestate I line. It was designed for use with standard 1960A or 1960B 4×12 cabinets and can drive a full stack. It’s not all solid state, since the Boost channel preamp uses a 12AX7 tube.

With two 100 watt channels the 8200 can create some very rich, wide sounds using its stereo chorus and stereo reverb. The Bi-Chorus feature gives separate rate and depth controls for each channel plus two manually selectable chorus modes. Other features include a foot switch input (for channel and chorus switching); stereo line out jacks which facilitate direct injection to P.A. or recording equipment; two sets (L & R) of twin speaker outputs for extension cabinet connection options; and two effect loops – a mono series loop and a stereo/mono parallel loop (mono send, L & R returns).


Trying it out on our bench confirmed that it did not work, and was producing no sound and no lighted indicators, including the power switch on/off indicator lamps. Thus it was time to disconnect the power and open up the cabinet by removing the top and side screws.

When an amp has no sound and no lighted indicators it immediately suggests checking the power supply, sometimes resulting in very simple fixes. With the on/off power lights not working, it was worth a quick check to see if the power (mains) switch was bad. Checking the input and output sides of the switch with the amp powered up confirmed that the switch operated properly. The internal lamps are just not working, which in our experience is not uncommon with the Marshall power switches.


With the easy and simple power switch possibility out of the way, we again unplugged the amp began the troubleshooting process with our absolute number one tip for electronics repair – a careful visual inspection of all components. In our experience, many electronics failures can be spotted visually. Using a flashlight (the white light LED ones are great for this) we shine a beam on every component, looking for any discoloration or something amiss. Discoloration in electronic components generally results from overheating, and in all but the mildest cases indicates that the component is bad. Note however that it does not indicate whether the component was the initial problem. The discolored component may have simply been the victim of a failure elsewhere in the circuit.

When troubleshooting we think of ourselves as detectives. We’re looking for “who done it” but know that we’ll need to gather clues to uncover the whole story. Not all problems leave a visual indication, but enough do that it is well worth the time to look things over very carefully.

During the inspection of the Valvestate we quickly spotted what appeared to be a brown, burnt resistor R2 (Photo 1) on the Reverb/PSU (power supply unit) board.

Photo 1 – Visual Inspection Found Burnt Resistor R2

Lifting a ribbon cable connecting to the same board revealed two more burnt resistors, R5 and R6 (Photo 2).

Photo 2 – Two More Burnt Resistors R5 and R6

No other anomalies were observed within the amp. Given that these resistors are part of a power supply board, we thought it unlikely that the parts had just failed on their own, but rather that they were likely damaged by too high current drawn elsewhere from within the amp. But doing any further diagnosis at this point would require a schematic, so it was time to find one on the internet.


A search for an 8200 schematic in Google found several sites with unanswered requests from other owners. However we did find a suggestion to try using the schematic for a Marshall 8280, a nearly identical but 80 watt version of the amp. This seemed like a reasonable ideas as the 8200 and 8280 are identical in features and front panel layout except for different power out levels. We found the schematic on on the site.


Having the schematic leads us to Tip #2: Thoroughly study the schematic to get to know the amp and circuits. You may not have the electronics knowledge to understand how every last circuit or component functions, but you need to be able to understand the basic amp systems (power supply, preamp, output power, etc.) and where they are located, how they are connected so you can work around the amp.

A review of the 8280 schematic (partial view shown below, with highlighted components) for the Reverb/PSU board showed that the two burnt resistors, R5/R6, are part of the +/- 12 volt power supply, on the inputs of the voltage regulator ICs. Finding both of these resistors burnt led to our hypothesis that something in the amplifier led to too much current being pulled from the power supply, causing the two resistors to overheat. Our hope was the voltage regulator ICs were undamaged, which is the job of the two 250 mA fuses on the input of the bridge rectifier DB1. We checked and found that both of these fuses were blown (open), which supported the idea of current overdraw.

The +/- 12 volt supply is one of the two main power supplies in the amp. It feeds all of the lower power circuits, reverb, preamp, chorus and so on, so it’s failure will definitely result in a no-sound amp. The other power supply is is B+/B- voltage for the output high-power section of the amp. We measured this supply at +/- 45.4 volts. Although our schematic doesn’t indicate what the voltage should be, a level of 45 volts seems about right for a solid-state power amp, so at this point we assumed the B+/B- power to be good.

Further examination of the schematic showed that the other burnt resistor, R2, is in series with the ring side of the JS3 jack, the “Left Parallel Return.” This is the Return input to the left side channel of the amplifier. As an input jack, this struck us as an unusual place to find a burnt resistor, since this would typically be a low-level signal part of the circuit. One idea we had was that possibly the amp’s output jack had been mistakenly connected to this return input, which would have led to a feedback loop and possibly the over-current condition mentioned above.


Our plan at this point was to repair the power supply and then power it up to see if there is still a current overdraw. If so, we’ll isolate sections of the amp to identify where the overdraw originates.

The two blown fuses are PC-mount fuses that we didn’t have on hand, so we ordered replacements from The Fuse Company. To keep working on the amp while waiting for the parts, we temporarily installed inline fuse holders with 250mA fuses. Replacing the burnt resistors R5 and R6 and powering up the amplifier showed that the power supply worked fine, but the resistors were beginning to overheat again. By disconnecting some of the inter-board ribbon cables we were able to disconnect most of the amplifier sections from this power supply, but the resistors continued to overheat. That meant that the overdraw was on the same board as the power supply, which indicated the voltage regulators or the reverb section. We cut some PC board traces to isolate the voltage regulator section and determined that the overdraw was coming from the reverb section. Since both rails (+/-) of the power supply were overdrawn we surmised (guessed) that the problem might be the MC1458P opamp, since it is fed by both rails. Removing this part ended the overdraw problem, and replacing the MC1458P with a new one restored the power supply to operating fine with all sections now powered up. By the way, our source for replacement MC1458Ps was


With the preamp now powered up we connected the amp to a speaker cabinet and found that one channel was producing a loud hum, while other had no sound output at all.

On this amp the heatsink and power amp board need to be removed together for access to their components. This gave us our first look at the underside of the board, which delivered some bad news – someone else had already done some work on this amp and it didn’t look terribly professional. The previous owner claimed that he was the original owner and had never had it repaired, so imagine that – we were lied to! “Caveat emptor,” (buyer beware) as the ancient Greeks used to say. From the solder pads and flux marks on the underside of the output power board it appeared that a good number of the transistors on this board had been re-soldered after the amp left the factory. Perhaps they had been removed for testing or perhaps they were replaced.

We decided to begin by tracing an audio signal through the amp. With our audio signal generator feeding a 1kHz signal to the return jack, we used our oscilloscope to trace the audio signal on the board and found almost nothing working. So we switched to DC testing, using our voltmeter to start reading the voltage on each transistor and quickly found some anomalies. For example, we found very different voltages on either side of the .33 ohm resistor R41, indicating the resistor was an open circuit. Removing it from the board confirmed that it was open. Further testing found different voltages on component leads that the schematic showed connected together, indicating that they really weren’t. As we dug into this problem it appeared that the soldering job by the previous technician had left behind some damaged PC board traces and/or cold solder joints. We used our de-soldering tool to remove the solder globs, and used our lacquer remover tool to reveal the bare copper traces. We repaired the damaged traces using light-gauge solid wire. In our experience, though it isn’t pretty, this approach works well.

While continuing with our DC testing we also noticed that top of the PC board for TR1/TR3 mis-identifies the emitter and collector. This may be true for other transistors, but we didn’t check. This mis-identifying created some moments of confusion for us. It’s just something to be aware of when doing repair work – PC boards sometimes have identification errors.

With the traces repaired and the DC voltages beginning to make sense we reconnected the audio signal generator and found that we could now trace the audio signal for one channel all the way through to the output terminals, and could see that the signal was being amplified along the way. So we hooked up our speaker and confirmed that we were getting a good sounding audio tone. One channel fixed, one to go.

To repair the second channel we continued our DC voltage testing, focusing primarily on the transistors, and soon found that the TR12 MOSFET (2SK1058) associated with the open R41 resistor had nearly identical voltage on all three pins. The B+/B- supply in this amp is about +/- 45.5 volts, and TR12 measured 45.5 volts on both source and drain, and 44.7 volts on the gate. Clearly this MOSFET had an internal short circuit. By removing it from the circuit we found that the remaining DC voltages appeared normal, and in fact this channel of the amplifier now operated. By operated we mean that it produced sound, although half of the signal waveform was missing and sounded distorted, due to the removal of TR12.

We ordered the replacement .33 ohm, 5 watt resistors from and the MOSFET from We also had to order a replacement thermal insulator to install the replacement TR12.

As shown in the photo, with TR12 removed you can see that the thermal insulator layer is scraped and damaged. The purpose of the thermal insulator is to provide better thermal conductivity between the MOSFET and the heatsink, while providing electrical insulation. Without electrical insulation, one lead of the MOSFET would be connected to the heatsink, and therefore grounded, which would not be good. Marshall used a rubberized thermal insulator strip, shown in yellowish tan in the photo above. This had the advantage of being quick and easy to install in a manufacturing environment. However it is not as effective a thermal conductor as is a traditional mica insulator with thermal compound. Since the rubberized thermal strip is not available from Marshall anyway, we had two good reasons to use an alternative. We could not find a source for a mica insulator for the TO-3P package used by the MOSFET, but we did find an alternative from made of Thermalloy.

Installing the new MOSFET began with removing a section of the rubberized insulator with a straight razor blade.

We then applied a smooth layer of thermal compound to the rear surface of the Thermalloy insulator, and to the rear surface of the MOSFET, and installed everything in the printed circuit board and heat sink. We should say at this point that this can be a messy process. Another complication was that the Thermalloy material is about 1/8″ thick, so it pushes the MOSFET forward from the heatsink by that amount. That required us to bend the MOSFET leads in advance to make everything line up. The installed MOSFET is shown below.

We also installed the new .33 ohm resistor, and did some more DC testing. Finding everything in order, and with the repaired preamp/reverb section in the signal chain, we were now ready to test the complete amplifier, and found it to be in working order. Our final step was to run the amplifier for four hours with a 1kHz signal on the input and a dummy load on the output, to test the amp at its operating temperature.

Category: Kit Building 101, The Repair Bench

Comments (13)

Trackback URL

Sites That Link to this Post

  1. My Homepage | March 18, 2012
  1. avatar Casey4S says:

    Excellent start to finish trouble shooting guide. Very informative.

  2. avatar Ross says:

    Thanks for a clear and practical description of troubleshooting. Very helpful. While ancient Greeks might have said “caveat emptor”, it is in the Latin tongue. Alexander Haig brought “caveat” into English usage as a stand alone verb in the late 20th century.

  3. avatar guitarkitbuilder says:

    Thanks for the language factoid Ross. Always good to learn something new.

  4. avatar southboundsuarez says:

    Nice, informative article. I like that you mentioned running the amp over a period of time at operational temperature. I recently ha a similar experiene with the purchase of a solid state Hughs Ketnner that was too cheap to pass up. (S/S amps seemingly appear to be abundant , cheap and ez fixes)
    I had a friend whom was in a bind, with no appreciable budget and needing a amp to immediately gig. Rather uncomfortable loaning my personal equipment I offered to help get this amp going for the price of a complimantary pair of darlington transitors for the output. Unfortunately in haste and somewhat of a noob myself we didn’t spend the extra time to be fully diligent enough to run extended dynamic testing. unfortunately this could have saved further frustrations for my friend with his gig. In our haste to get him outa the shop and to rehersal we overlooked fully tightening our components to the heatsink and one the transistors wasn’t making good thermal conductivity and would overheat and would drop off in power…. The extra time spent being diligent and testing on the work bench would have saved much face!
    My question is what is your dummyload composed of and how much power can it continously dissipate? At what power levels do run your extended testing at?
    How about a article on basic test gear, the practical application and how to build and acquire some the basics such as your signal generator and especially a inexpensive, simple down and dirty dummy load?

  5. avatar Brad says:

    Hi. Oddly enough, i have this exact same amp and am just now looking for someone to repair it. Ironically enough, i have an Electronics Degree but don’t know where to start on this thing. You guys wouldn’t happen to be in Houston, Tx would you? Anyways, although very informative, i’m not sure rather or not this article will help me in my endeavors. So, how much did you guys end up spending on all of the parts for this? I don’t know rather to attempt to fix this or just look for something else. I loved this amp when it was working. Thanks

    • Dave, we are in Northern Virginia but not in the business of fixing customer amplifiers. Since you have an electronics degree, perhaps you could learn to troubleshoot the amp yourself. Do you have a digital multimeter? What are the amplifier symptoms?

      • avatar Brad says:

        Name’s Brad, not Dave, and i’ve started troubleshooting. No visible problems with any internal components. The amp went through a minor fire years ago. Hoping, as it appears, that all damage is cosmetic. I did have to bypass the power rocker switch because the front plastic melted. But, the amp does seem to power on fine. For all i know, the thing still works, however, i don’t have the means to test it until later this week. So, no symptoms just yet. I do have a digital mulitmeter and i’m getting 120VAC to the amp. Five of the pots, including the master volume, are ‘stuck’. So, as it stands, i’ll probably start by replacing those and the rocker switch. Thanks

  6. avatar Dennis L. Ellrod says:

    Can I get some suggestions on troubleshooting a B-52 LG100a solid state amp?
    The amp has a very low sound output on the clean channel. The sound goes up and down with the volume pot but it doesn’t come close to the 100 watts of power. The distortion channel only makes low static noise.

    One more request. Where can I find a good repair manual for Solid State amps? I have searched all over the inter-net and can’t find a repair manual that has detailed trouble shooting for a Solid State amp.
    Thank You, excellent site, Dennis my e-mail

  7. avatar John Simms says:

    Any idea what caused this damage? Anyway I’d love to own one of these amps.

  8. avatar Will Smith says:

    I have the bi- chorus combo. I had experienced a power drop, so I replaced the tube. Much to my surprise that didn’t work. I read the above post and looked over the circuit board for anything that looked burnt out. I found nothing. I noticed that the reverb box had springs lying at the bottom of the amp. I unplugged the reverb cords (red and black) and put duct tape over the loose rc jacks. I put the amp back together. I still experience a fuzzy power drop. I tired turing the effects knob all the way down. Bingo- It worked!!!!! I hope this helps someone.

  9. avatar Charlie says:

    Hello I have this exact same amp and it means a lot to me since my father give it to me as a gift, I have the challenge to restore to a factory condition, could you please help me on recommend me with some one that could repair it, I live in Mexico City but I could send it to any part of the world just to get that factory condition

  10. avatar Edu says:


    I have found a Valvestate 8280 in poor condition. I have rebuild but the clean channel don’t sound. Maybe at full volume sounds some sound, but depreciable. I switch to overdrive channels and the two works great. I have replaced the volume pot and the same problem (in the schematic shows 1M pot, but in the pcb shows 220k and have installed 220k. In the schematic shows 220k tone pot, but it have 1M and the same in pcb).

    Can you help me? the problem can be in TR3? it is possible to fail this zener diodes?


Leave a Reply

nine × = 36