marshall valvestate VS265 problem

[g1]

Saw another old post about a VS100 that mentioned it having 12volt regulators. As JazzP stated, that schematic also shows 12V regulators but shows +/- 15V so the 12V regs. may well be stock. You said one of the resistors had 41V at one end and 25V at the other. Thats a 16V drop across what should be 180R. 16/180 = .089 amps, times 16V gives approx. 1.4 watts. The resistor is shown as a 3 watt so it is not really overheating (assuming it is actually 180 ohm.)

[andrewkirsanow]

I think they maybe did change the spec as all of the schematics of the larger valvestates seem to specify 15V rails, but the circuit diagram specifies the use of 12 volt regulators. My VS65 however, specifies 15V rails but uses 15V zener diodes in the PSU to regulate the rails at 15Volts. So, I think that maybe the diagrams were created, then altered to use the slightly lower rail voltages, which wouldn't really affect the design a great deal, it's still plenty of headroom for small signal op amps.

The only real way to verify whether the heat is 'normal' would be to measure the voltage across each of the two resistors with your meter. When you have those readings, you can get a value of each of the resistors R20 and 21 by unplugging CON3 (when the amp is off and unplugged), and measuring the value with the ohms range on your meter. Removing CON3 first ensures that nothing else will be connected in parallel with either resistor, so you will get a true reading (at least that's the way it looks on the diagram, you'd have to verify that one end of each resistor connects to a pin on the connector and nowhere else, otherwise, unsolder one end of each resistor).

You can then use each resistor's value, along with the voltage dropped across it to calculate the power dissipated in that resistor. If it works out well under the 3W rating for the resistor, then it's running within it's specification.

It really depends how far you want to go into this. How hot would you say the regulators are getting? Are they far too hot to touch? The exact part code on the regulator would be helpful in determining the maximum allowed current etc. which could help you to see if they're being worked too hard. My gut feeling is that they should be only operating well under their rated specs, as they have no heatsinks on them at all - the standard 7812 can only handle a bit under 2.5Watts with no heatsink.

[rockjunky]

Valvestate VS265 Blown! - Marshall Amp Forum

just saw this thread, same amp different looking resistors .....could it be possible there resistors are a lower wattage hence more heat? it shows 3W on the ones in the thread but not on my friends
my friend said these resistors looked like they had been resoldered to


I removed CON2 and got about 189 ohms for both of them without taking them out of the circuit
I removed the salastic and its shaky as! will definitely have to put something on the legs

I will check the drop across them , current etc right now


andrew the REG's are to hot to touch same as the resistors

[rockjunky]

voltage on R20 supply -40VDC , other side is -28VDC = 12V drop 12/180=0.066 X 12= 0.8W?
voltage on R21 supply 40VDC , other side is 23VDC = 17V drop 17/180=0.0944 X 17 = 1.6W?

I used 180 for OHM's but when I measured it was more like 189 ....I probably should take them out and check

needless to say they are VERY hot...I wonder if they might be 1W that got replaced instead? could this be the heat problem?

the REGs are very hot to

these had pretty bad solders where the traces lifted....they aren't fantastic (now that the salastic is gone they are pretty flimsy) but the solders are ok..its just the trace is lifted a bit

[g1]

If they were 1 watt at least one of them would likely have burned up by now. It is possible they are only 2 watt, hard to tell as new type resistors can handle more wattage in a smaller package. They are supposed to get hot, they are converting excess voltage into heat. The amount of heat generated has nothing to do with the wattage of the resistor, a five watt resistor in that circuit would generate the same amount of heat, although it may feel cooler as it has a larger surface area.
Taking them out to measure them would be a good idea, or at least disconnect one end. If they are off by much and you have to replace them you could use 5 watt resistors if you prefer. The regulators also convert excess voltage into heat so they will run hot. If they are really overheating the 12V output will probably start to drop off.

[andrewkirsanow]

If you've double checked and one end of each resistor only goes to one of CON3s pins, and nowhere else, then removing that connector will effectively isolate that end of each resistor, so the readings will be accurate for the resistance measurement. 189ohms seems reasonable(ish) given that the gold band states that these resistors have a tolerance of 5%, so they are at the upper end of their value range at 189ohm. However, that may suggest that they are suffering somewhat, perhaps decaying with the heat - you may expect values from 171 to 189, so it's suggestive that both read exactly 189 Ohms? On the other hand, those calculations suggest that the resistors are carrying only 66 milliamps and 94 milliamps respectively, which doesn't seem like a great deal. I suppose you could upgrade those resistors to 5W types to play it safe, after all, I'm not too sure on sight that those are 3W, they could be, but I'm not sure.

I'd also double check the components in the main PSU section to make sure no AC component could be getting through to the regulators. Particularly the rectifier diodes and the capacitors. This sort of testing is easiest with an oscilloscope, but you can check the rectifier diodes in circuit with the power off, as a short in one of those would show up readily.

Now then, what I would do next is I'd use my meter on a current range and remove the output leg of one of the regulators, then clip the meter leads in series with the regulator, between the removed output leg, and the solder trace it used to be connected to. Then I'd measure the current drain on that regulator. I'd then repeat this on the other regulator and get an idea of the current demands on each regulator. This would let us know for definite whether the regulators are being overworked. However, you should only attempt operations of this type if you have clips to firmly attach the meter leads into the circuit, and are confident to do this type of in-circuit measurement. And if you do it, remember to put the meter's positive lead back in the Volt/ohms plug afterwards as forgetfully leaving it in the meter plug used for current measurements can be very dangerous when trying to measure a voltage later on!!

If those regs are working within their capable range, and just running hot due to bad design, I'd consider moving them onto a heatsink, or redesigning the power section personally!

[rockjunky]

If you've double checked and one end of each resistor only goes to one of CON3s pins, and nowhere else, then removing that connector will effectively isolate that end of each resistor, so the readings will be accurate for the resistance measurement. 189ohms seems reasonable(ish) given that the gold band states that these resistors have a tolerance of 5%, so they are at the upper end of their value range at 189ohm. However, that may suggest that they are suffering somewhat, perhaps decaying with the heat - you may expect values from 171 to 189, so it's suggestive that both read exactly 189 Ohms? On the other hand, those calculations suggest that the resistors are carrying only 66 milliamps and 94 milliamps respectively, which doesn't seem like a great deal. I suppose you could upgrade those resistors to 5W types to play it safe, after all, I'm not too sure on sight that those are 3W, they could be, but I'm not sure.

I'd also double check the components in the main PSU section to make sure no AC component could be getting through to the regulators. Particularly the rectifier diodes and the capacitors. This sort of testing is easiest with an oscilloscope, but you can check the rectifier diodes in circuit with the power off, as a short in one of those would show up readily.

Now then, what I would do next is I'd use my meter on a current range and remove the output leg of one of the regulators, then clip the meter leads in series with the regulator, between the removed output leg, and the solder trace it used to be connected to. Then I'd measure the current drain on that regulator. I'd then repeat this on the other regulator and get an idea of the current demands on each regulator. This would let us know for definite whether the regulators are being overworked. However, you should only attempt operations of this type if you have clips to firmly attach the meter leads into the circuit, and are confident to do this type of in-circuit measurement. And if you do it, remember to put the meter's positive lead back in the Volt/ohms plug afterwards as forgetfully leaving it in the meter plug used for current measurements can be very dangerous when trying to measure a voltage later on!!

If those regs are working within their capable range, and just running hot due to bad design, I'd consider moving them onto a heatsink, or redesigning the power section personally!

yup I got you on the current!

I dont have anything clip wise to do that thou sadly....

I was thinking it wcould be something in the main PSU.....
tonight I tried a 8240 valvestate and there is a very similar circuit and the resistors there were not even hot touch, the REGs were pretty hot thou
measured across those resistors it was +/-21VDC and on the other side was +/- 19VDC ...quite a bit different but it is a different amp to.....

I think my friend is keen to give it a bit of a test and see how it goes....maybe it was runnign like this heat wise before? its hard to say :/

I'll keep you updateds thanks for helping so much!

[andrewkirsanow]

Sorry if I came across as patronising, but I was surprised to learn how many people don't know how to use a multimeter to measure current flow, so I wanted to be on the safe side, just in case!

Looked at the schematic for the 8240, it seems that the transformer secondary on that one runs at just about 20V or so, so there wouldn't need to be the same voltage drop across the resistors. In the case of the 65Watt, the power amp section requires +/-40V or so, and to save on using a transformer with another winding, Marshall decided to resistively drop a lot of that voltage, and then use voltage regulators to obtain the final 12V rails for the rest of the pre-amp. Given the amount I'm seeing with blown out regs/resistors (such as the one you linked earlier in this thread), I'm starting to wonder if the design is a bit flawed, and the regulators run very hot in normal use - if that's the case, then I'm disgusted with a company like Marshall as these amps weren't exactly inexpensive from new. This is a recipe for disaster in these closed, badly vented casings - at least they could have screwed the damn things to the chassis! If your friend tests it out, and everything seems to work as it should, we can only assume bad design. If it goes 'pop', and you get it back in to look at, then we'd have to start to consider the possibility that a failing component somewhere is imposing an excessive load on the PSU section.

Good luck anyway!
Andy.

[km6xz]

Andy, the problem is common, where designers complete a product than it is localized by others. New versions with details changed due to laws, certification requirements, voltage and frequency sources etc. This happens a lot, where the unit works perfectly in one country but is a warranty problem in another.
There is every indication that the people who designed the pre-amp section expected the rails to be +/- 15v on this one. All the components, caps, ICs etc are rated for tolerances fitting for a 15volt rail. The transformers for the design market might have been right, or assumed the correct mains voltage, but the spec for the 120 volt version might have come in assuming a lower voltage. This has happened in hundreds of models over the years.
When I got those units in, I lowered the stress by installing 15 volt regulators, everything ran cooler and nothing de-soldered itself( a very common warranty problem). The resistors are running too close to their max rating as it is, considering the ambient air temp is higher than the 20C they are rated at, and those who put glue or hot tack on the resistors to keep them from tearing up the pads further reduce their heat dissipation. The 240 volt versions of the amp do not have this problem. The regulator IC are not recommended to be used without sinks, and no specs are guaranteed unless one is used. The part needs to be greatly derated even in free air, let alone in a warm amp with no ventilation. The part maker does not even assume responsibility for specs if a heat sink is used without their design heat transfer function.
We have the choice of leaving it as it came and risk long term reliability problems or make changes to compensate for reality rather than assume the designers had a close connection with the localization process. Tell the customer and let him choose but note it on the work order if they accept the risk, so they can't expect a shop warranty repair when the parts melt down. Can anyone cite a downside to adjusting the rail voltage to what was intended and the resulting in the power supply components running as intended? None, sound is helped, reliability is helped, and longevity is helped.
Cooler equipment is cool, it last longer and has fewer problems.

[Steve Conner]

Yes, if you look back through all those old forum posts, you'll see a lot of complaints about these dropper resistors in SS amps unsoldering themselves. Valvestates, Fenders, SWR bass amps, etc.

I saw one where the guy replaced the dropper resistors with the gold aluminium housed type, extended on wires and screwed to the chassis with big gobs of thermal grease. It was a hobbyist fixing his own amp: too much time and effort for a pro repair shop.

It's often not possible to adjust the rail voltage. Sometimes the transformer just has one "120V" winding with no taps. Replacing 12V regs for 15V ones is a good trick, but if they're already 15, what do you do?

[andrewkirsanow]

My VS65 already runs at +/-15V rails, for some reason they skipped the regulators altogether and used zeners to regulate at 15V. I can't see any possible reason why they opted for 12v instead of 15V. Although I think a 24V range is more than ample headroom for this sort of pre-amp, I agree with all of your other reasons in favour of using the 15V regs, such as reduced heat dissipation etc Unless they found that the 15V rails were running some of the other ICs too close to max spec? I haven't looked at the supply specs for those op amps personally. I think if I had one of these, as I stated earlier, I'd modify the PSU. I'd love to speak to the designer of these at the time, to answer the 12V mystery!!!

[km6xz]

A bucking coil in the secondary if the power amp rails are right. A bucking coil in the primary if the whole primary winding need to come down. This works if the core is available after taking bells off. A few turns looped though the core in the same magnetic field, connected out of phase with the primary will lower the output voltage and lower primary current. Everything runs cooler.

[rockjunky]

seems to be running fine now! thanks so much guys, great knowledge here!

[andrewkirsanow]

Really pleased to hear that dude! And this has enlightened me too, I was thinking of getting one of those, or a bigger one to upgrade my VS65R, which is a little underpowered and 'weak' for a lot of venues. If I do obtain one, the first thing I'll be doing is opening it up and soldering over the board, then upgrading the power supply!!