Marshall 100W Mains Tests Data

                Marshall 100W Mains Tests Data

I was a bit thrown by the low DC resistances of the windings of this at first, as 2-5 ohms seems really low compared to the smaller Danbury transformers I have measured (30 ohms primary, or so), so I removed it to test it properly using a Variac. I initially assumed it had a short somewhere due to these low readings.

I want to clean the chassis mount point grounds anyway. This was interesting to see the 3 different secondary windings outputs work in terms of their turn’s ratios also.

The peak to peak output potential of this transformer is above 1000V. This would account for the under load figures I have read about being 400-600V DC for A+ from each ½ cycle. I want to ease this transformer up in small voltage stages also, as it has not been electrified for 30 years, so the insulation could break down at any moment. I went to 120V RMS for this session only. I can run the amp at this ½ UK voltage later if necessary with the Variac, as a working minimum if all goes well, knowing that it is OK for that level.

Removing this also made life easier to remove the rectifier diode housing which looks like the faulty component at this stage, now the transformer seems ok for all windings.


Quick reference page for this powerful Marshall 100W transformer showing the general linearity, and the phase of orange and green is the same:

240V winding     
Vin RMS Meter        Mains Orange/Purple        Secondary Green/Black

            10V            30            50

            20V            60            100

            30V            90            150

            40V          120            200

            50V         Off scale        250

            60V         Off scale        300

            100V        Off scale        500

220V winding    
Vin RMS Meter        Mains Orange/Green        Secondary Green/Black    

10            30            50

200V winding    
Vin RMS Meter        Mains Orange/Red        Secondary Green/Black    

            10            30            50

            40           120           250

120V winding
Vin RMS Meter        Mains Orange/ Brown        Secondary Green/Black

            10            30            100

            20            60            200

            30            90            300

110V winding
Vin RMS Meter        Mains Orange/Pink        Secondary Green/Black

            10            30            100

            20            60            200

            30            90            320

240V winding
Vin RMS Meter        Mains Orange/Purple        Secondary Heater Ctr-1

            20            60            0.82

            40            120           1.6

            50            150           2.0

240V winding
Vin RMS Meter        Mains Orange/Brown        50 OHM winding- 27k/diode

            20            60            25

            30            90            35

            40           120            50

            70         Off scale        90

            90        Off scale        120

            100       Off scale        130


Since this data was logged I have checked the rectifier and rest of the board to a point where I think there is nothing actually wrong with this amp in the first place. I guess a valve blew 30 years ago, and it was thought there was something more serious? People and common sense…

The first test with the Variac to mains 240V level has been done with no smoke or incident, showing an anode voltage of 523V below:

This DC reading was taken at the Standby switch to chassis ground:

I moved the Standby switch to after the diodes as with the other amps so the caps don’t charge until this is on. The original Marshall 6.3V bulb was blown, so I used a 240V neon across the mains primary switch to Ground. There is no hole for a Standby light – which personally, I think all amps should have. Just for interest’s sake, I took the Variac opportunity to push these 220V AC neons to their limit – they do stand 525V DC or more across them!

The next step is to do the same test with all valves in then do an FLS signal check etc.

I will have to build a frame first, as this amp takes about 10 mins to discharge DC down to about 60V DC, with the valves out, and ALL the large caps need to be checked separately to be sure they are drained before working on it, so I want it stable on the bench so I can see the valves heating up and know the meter connections are visible too if possible. I can’t afford a slip at these voltages for sure.

With the valves in, this test was a different story – there was a fault somewhere causing the DC level to match the AC input levels only – 240V or so. No smoke or burning though so just have to find the component that seems to bypass the yellow DC cans now – that’s my current theory anyway.

I thought an approach would be to remove the valves 1 by 1 until I get high voltage DC at the rails again, to narrow down the area section by section.


I decided to check the filter caps again as I was getting full wave rectification, but no DC filtering on the scope, so I de-soldered each yellow and blue can, and checked them with the cap tester in turn. It was a bad ground joint on the blue filter can that prevented the filtering and HT rail separation. Once re-soldered, I powered up again with no valves, and got the double mains value of 523V DC again. Once de-powered and drained, I put the valves back (I even checked all these in the Ashton again to be sure!), attached the speaker output to the 66W Marshall Marquee speaker from the Fallen Angel, and I got the 523V also this time, so tentatively wound up the Channel 1 volume to hear the comforting sound of mains hum through the speaker – the first time that’s been welcome! It means there is no OT problem. I turned that down and tried Channel 2 volume, which was not as mains noisy, but worked also.

I tried all the tone and presence controls, which work too. I tried a test signal from FLS, with the 10 X Probe on the speaker output. This looked nicely symmetrical, so happy with that. I decided to plug the guitar in the low input of the less hummy Channel 2 and heard a nice, but slightly muffled tone. The treble sorted that out, so I twanged a bit for 10 mins, thinking “this amp hasn’t been heard for 30 years – what a waste!”
This amp discharges to about 10V in only about 30 seconds with all the valves present, but NOT if the valves are removed – the large can caps stay charged for a long time.

Just a few things to log and fix now, then I will re-house it, and try and get it to college, or Andy at Modern Music in Truro:

who said he would try to borrow a 100W cab to hear it cranked through…

If all goes well there will be a video of it all.

As it will probably be sold now, if anyone is interested, email me:

and I will pass offers on to John (sensible offers ONLY please from collectors or gigging musicians who know what this is, and that it will need some TLC to become a workhorse again. Possibly new pots, bias slider, re-biasing, re-sleeving of transformer wire insulation and all electrolytics changed out and/or reformed, though re-forming looks like a bad idea as they will fail sooner or later and may cause serious damage to the mains transformer if they fail short:

For now I will keep it as original as it is, maybe looking at fixing the Channel 1 mains noise, and see what else happens once I can get it somewhere to crank it up a bit, and hope nothing breaks down again.
(Turns out the noise was probably the neon light flickering under probable duress at 523V DC.
I can’t face having this back on the bench! I want to see John’s 1960s Fender Tremolux next!