Geloso G1/1040 Amp repair

Burns, Selmer, WEM, and more

Re: Geloso G1/1040 Amp repair

Postby zjokka on Tue Nov 01, 2016 1:18 pm

trobbins wrote:Are the heater voltages ok that are powered from the bias?

What would happen if the heater voltages of any tubes were completely off?
trobbins wrote:PS. 31V bias reading indicates 39V from bias supply. Nominal 150mA through heaters would require 12.6+12.6+100x0.15 = 40V.

If you are supposing the bias for the power tubes is -31V, why were you asking above? I have -31VDC coming off the rectifier after replacing that 500uF capacitor.
Best to insert cathode current sense resistor for each EL34, then you can easily assess if each valve is ok, and whether you have nominal balanced valves,

Maybe I should just jam in a couple of new, balanced EL34's from my stash? Or is this too easy?
and can calculate valve dissipation to see whether the bias voltage needs to be tweaked up/down a bit by changing the 8K2.

Something's missing here -- you want me to install biasing resistors to
trobbins wrote:You may want to disconnect the feedback whilst setting up the bias - just in case there is some oscillation happening in the background that you aren't aware of.

Yeah, I just may...

zjokka
Peasant
 
Posts: 13
Joined: Sun Sep 11, 2016 3:37 pm

Re: Geloso G1/1040 Amp repair

Postby trobbins on Wed Nov 02, 2016 10:32 pm

If you have 31V at the bridge rectifier output, then you likely have 8.2/10.4 x 31V = 24V as a bias voltage for the output tubes - due to the extra filter circuit shown in the schematic. That seems too low.

31V also seems a bit low for the supply of input stage heater current, as the heater voltages are likely to be significantly less than 12.6V - 10% = 11.3V each.

So I think you still have a problem that hasn't yet been identified. Perhaps if you remove the first heater/valve, so there is no heater loading, and check what happens to the 31V level, and also to measure the power transformer winding AC voltage for that bias supply.

If an amp is all ok, then putting in new valves is usually no issue. If the amp is suspect, then putting in new valves may just stress or damage them.

Some people use the bias measuring adaptors to confirm each output valve is appropriately passing the right idle current, with the right idle voltage, to cause an appropriate plate dissipation. Adding cathode current sense resistors is another easy method to confirm what is happening. There other harder methods. And then there is the easiest method of doing nothing. Your choice - just a suggestion, and advise that any repair normally has to check all areas of operation, in case other problems also exist - its an old amp.
trobbins
Court Jester
 
Posts: 74
Joined: Wed Oct 14, 2009 7:30 am
Location: Melbourne, OZ

Re: Geloso G1/1040 Amp repair

Postby Racing on Thu Nov 03, 2016 7:22 am

If memory serves me you told you live in Belgium?
In that case you´re supposed to se just shy of the 240VAC mark at the wall and that renders that in 220VAC mode,the amp that is,you should see slightly elevated B+ numbers.
That also bring that the bias voltage should respond in a similar manner.

B+ is given at 340VDC for these if memory´s with me,and in 220VAC mode that should bring along the lines of 360-370VDC approx,under load and in idle. IOW that bias number should be approx minus 33 or so coming off the rectifier.
Mind you,this depends greatly on what´s seen at the wall.(read-measure wall voltage and know)

In short it seems you´re about on mark,and this is a rather far cry from the numbers given a few posts back. For the better :thumbsup: .

You also mentioned that you´ve figured the OT out. Good. If you look at how the OT´s wound on a schematic i´d say it´ll become even more clear. Please don´t forget that the OT secondary always needs some sort of ground reference to work properly. This with or withinout negative feedback in loop. Seing we´re working with AC at that point it really doesn´t affect all that much where this ground reference point is made.

Trobbins suggestion of installing 1 Ohm/1w resistors between powertube socket and chassis ground is valid. It is a cheap and effective,not to mention way more safe,way to quickly see if things are in order with the powertubes and its surroundings.
As the resistor is of 1 Ohm the reading taken across the resistor can be directly translated from mV to mA and thus inserted into Ohms second law together with B+ to get a reasonable reading of anode loss in idle.
I for one at least pull this stunt whenever i feel the need,which is basically always. This way there´s always a convenient measuring point.

Haven´t read up that far back,if you already have or haven´t, but to make the entire bias setup adjustable what you do is that you replace the 8,2k resistor in the bias circuit with a 22k trimpot. This way dialing bias in becomes a breeze.

One of the advantages of using the 12AX7/ECC-83 and similar tubes is that the offer leeway as far as heater voltage. This can either be AC or DC but atop that it can be either 6,3V or 12,6. IOW when it comes to measured heater voltage,just add up and be done with it. Hooked in "12 volt mode" the current digested by them tubes becomes half of what would be in 6,3V mode. IOW a total of 300mA.

Provided fair signal the 1040 becomes a rather vivid and powerful amp,and in turn one that lends itself to modification as you see fit and needed.
Racing
Squire
 
Posts: 749
Joined: Wed Dec 21, 2011 11:07 am

Re: Geloso G1/1040 Amp repair

Postby Racing on Thu Nov 03, 2016 7:46 am

Btw.
On that OT secondary ground reference. WHERE this is taken becomes an issue first as negative feedback is introduced. If you read the "setup manual" at the rear of the amp and take a look at the schematic,the OT,at the same time i bet it´ll all become clear to you.
For an OT hookup,yes the "Geloso way" is complex. OTOH it also hands a downright myriad of options..so fair trade IMO.

FWIW the older amps (read,-40´s and early -50´s) lack this and the options presented are way more "normal" to the number. This is easily spotted as these amps have a mere 6 points for hookup at the rear.

FWIW,and if you´re into it,try picking one of the earlier Gelosos up. Cool looking,at least IMO,and performs just as well-and then some-as the later offerings.
Right now for instance i´m putting together a G.30 (1946) a G.211A(1953) and shortly a G.229 (1953 as well) to become guitar amps instead.

Image

To give you an idea here´s a mid -50´s G.225
Being reset for its new purpose in life...one serious contender as far as rocknroll amp.

Image

Here in turn a "mystery" G.60 As you can see a slightly larger amp. Again focused on RnR.

Image

Then a G.213
These are about the size of a 3215,just running twin 6V6´s..which of course adds a different character vs the EL-84 amps.
Racing
Squire
 
Posts: 749
Joined: Wed Dec 21, 2011 11:07 am

Re: Geloso G1/1040 Amp repair

Postby zjokka on Sat Nov 05, 2016 8:40 pm

Yes, I live in Belgium, we have about 230V in the socket... but hard to put DMM in the wall socket.


Here's an update: most important things is that, measuring the voltage at first 40µF filter cap, I got like 245-250V, but when applying more pressure to the solder point, it jumped to 340V. Then I'm getting about 343VDC to the plates of the EL34's. But when I had this desired (right?) high voltage on the capacitor, the amp did not work. Touching the input gave no buzz. When the amp shows 245V here, it does work. Anyway, after finding the problem with the solder joint (I thought), I turned off amp, drained caps, heated up the iron, reflowed solder, checked for continuity between cap and HV rectifier, found it ok, fired it up again. Some more measurements showed that it still jumped from 245V to 340V when wiggled, so I concluded upon a mechanical defect inside the capacitor. So replaced the 40µF+40µF can with 2 pairs of paralleled 22µF/350V's...

Oh yeah: one side of the 40µF side was dead as Zed; the other side measured out perfectly... It goes to show you only know when you measure them properly.

Amp is now back to how it was -- working -- never seeing 340V

Did do some bias checks according to above mentioned method: which is not ideal...
Code: Select all
           resistance   v drop   plate current   plate voltage   plate dissipation
EL34 (1)   75,9 ‎Ω   1,85 V   24,37 µA            245 V            5,97 W
EL34 (2)   85,2 ‎Ω   2,35 V   27,58 µA            240 V            6,62 W


I will install the 1Ω biasing resitors - but need to head to the store for this and other stuff.

But this is with the old, original tubes, maybe I should put it new tubes?

About the output transformer: I know one speaker lead is always ground, because I took this as an absolute, I had the speakers hooked up wrong. There is a chassis screw exactly besides the number 2, so I hooked up to that -- which didn't work. Now I have lugs 1 and 4 connected and speakers hooked up to 2 and 3. Which is how I read the Geloso chart... But indeed there is no connection to chassis ground now!

Only now did I notice the previous work done on this amp: some resistors, a.o. the biasing 8k2, was replaced by a modern carbon comp (in green below). But there are others too (in blue), so I need to check what went down there. Strange that somebody replaced resistors with original values and leaving the capacitors in place... I am going to trace the circuit and mark the replaced resistors. Also, one of the 470‎Ω (in red) needs to be replaced; it has seen considerable heat stress. I don't understand why the biasing resistor was replaced but the parallel capacitor was kept. Will replace that too before doing further tests.

Image

BTW, wires coming to these main capacitors were red (hv) and white (coil), so these Italian fasionista's chose green for the ground wire. It's the little things about Geloso...

No need to get me hooked on Geloso, I live in Belgium, the place to buy Geloso amps much cheaper than anywhere. Two days ago, a nice G1110 120W with quad EL503 popped up for €80 untested. Maybe I should just buy it as a gamble for the EL503's? If I could convert to another tube like EL34 and sell the original tubes, i'd make a profit. There is also a 1070 for sale for €150, a 6L6 15watter for €200.... (I'd take it for €100 but have 3 Geloso 15W amps now)

More testing and results soon...
zjokka
Peasant
 
Posts: 13
Joined: Sun Sep 11, 2016 3:37 pm

Re: Geloso G1/1040 Amp repair

Postby trobbins on Sun Nov 06, 2016 3:02 am

With the replaced filter caps, you still seem to have a HV supply problem, as you should be measuring around 340VDC.

Can you measure the VAC of the HV secondary with the standby switch off (ie. across the bridge diode ac terminals) ?

Your bias measurements appears consistent for a low B+ of 240V (bias current is in mA not uA). With around 340VDC supply, and an appropriate bias voltage, your target should be to operate each valve at 20-25W anode dissipation.

The heat stressed 470 ohm screen stopper is typical - many restorations increase the power rating of that resistor (eg. to 1-2W in your case).

The anode lead off the EL34 socket should be routed as far from the grid wire as practical - the photo appears to show it crossing at right angles (good), but somewhat close (bad).

The feedback winding 0V connection should be made at the same spot as the 2k2 cathode resistor connects to 0V for the 12AT7 feedback triode.

There may be a lot of feedback being applied, as there are a few high frequency tailoring circuits being applied for stability - so any changes to parts / values around that output stage and feedback winding could easily cause instability - and a squarewave test with resistive load would be appropriate once the biases are all back to nominal.
trobbins
Court Jester
 
Posts: 74
Joined: Wed Oct 14, 2009 7:30 am
Location: Melbourne, OZ

Re: Geloso G1/1040 Amp repair

Postby Racing on Sun Nov 06, 2016 9:52 am

I agree,and certainly something still ain´t right.

I voice for the following,as we need to isolate the problem at hand.

According to the schematic standby works between bridge negative and chassis ground. First up,with amp off and drained,do a DC resistence check between chassis and bridge negative. This should show zero Ohms,or very close to it. If not,replace or delete that switch entirely.

Second up we need to isolate the powertstage per se. Simplest way to do this is to lift one leg of that first drop resistor in the voltage rail,which is of 10k. By lifting one leg of that we to the letter "kill" the rest of the amp which is exactly what we need as we want this piece up and running in proper form again.

Your choice of doubled 22µF/350VDC can/will come up short. These capacitors needs to be of 450VDC minimum as voltage will increase from referenced 340VDC first of all-seing mains voltage today-and what´s more it will jump if a powertube fails down the road too,making the margin of 350VDC caps come up short-making the amp unsafe.
Me i´d just throw some 450-500V axials at it and call it a day.

In turn measure for resistence between pin1&8 of the powertubes and chassis. This should read zero Ohms too. Pin 1 being g3,the surpressor grid,and pin 8 being the cathode.
Then measure each screen resistor for resistence,which should be around the 470 Ohm mark.
Trobbins is absolutely right in that we these days often up that value a bit first of all and what´s more up the wattage of those resistors as well. I´d use regular 1k/5w ceramics for this.

Now.
You´ve obviously got a simpler ESR meter. Good. You´ve taken measurments of the old capacitor of 50µF+50µF one and deemed that shot. Have replaced with doubled 22µF for the first two hits.
I take it these capacitors,the new ones,are fresh? Ie;new? They STILL need to be replaced with caps of higher voltage capacity.
In that case..

Start by reflowing every applicable solderjoint for the powertubes and PSU(power supply unit) in sight. I don´t care how fresh they might look,hit ´em with the soldering gun and a fresh dab of leaded solder and be done with it already.
Reason is simple.
We still have a problem. Hence why i ask you to take measurments per above. Something is amiss,and we KNOW this from schematic and experience and we need to find out why.
Sure. An oscilloscope will help but for these basic moves it isn´t called for really.
Make DAMN sure the fuses involved are of the correct rating. Now..there´s two "death caps" in there as well...... They are of 2,2nF and you see them in the schem on the powertransformers primary side. Cut these out and discard.
Install a 3 prong mains if not done already. Ie;phase,zero and safety ground. 3 wires all in all. Safety ground SHALL go to a point used SOLELY for this on the chassis. NOTHING else goes to that point. A regular modern mains wire for a computer or whatever along those lines will work just fine. If you are to keep the mains wire as a steady state install just cut the socket off and use the wire per se. In such a case be sure to reuse the stock mains wire "locks" so the wire can´t come out if yanked.

So. WIth everything is sight reflowed flick the switch again. Now take a measurment of AC inbound for the rectifier. Ditto for primary voltage inbound for the powertransformer. We need that referenced too.

Although you can very well measure bias as you have,via "the shunt method",it might still be an idea to install a 1 Ohm resistor of 1 Watt between each powertubes pin 1&8 and ground. Mainly as it´s safer to measure anode loss in idle that way.
B+ voltage should now be approx 340VDC+. If not we have an issue and need to figure out why.

As you can see from the schematic there´s a fuse for HT AC secondary inbound for bridge. Check this fuse AND it´s fuse holder for integrity. That means that the fuse should sit firmly in its holder and be free of dirt/oxides and what not. Voltage drop across that fuse should be ZERO VAC... Fuse holders can often be retensioned sligtly need be. If not,replace.
Check inbound AC voltage with standby OFF. We need to have that value referenced.

While at it check integrity,again with amp off and drained,for the bias circuit. Reference each resistor with your DMM. If memory serves me you already replaced the 100µF e-lyte in there? In that case good. If not,replace. Why i say replace instead of measure and check is that an e-lyte of that size is dirt cheap and..old is old. Period. Remember that this is NEGATIVE bias voltage why the e-lyte needs to be installed "the other way around". Ie;it´s positive lead goes to ground.
Further i still vote you replace that 8,2k resistor vs ground for the bias circuit with a 22k trimpot. It is a really cheap and sensible solution to any bias issues that might come up.
But..again. Check all involved resistors vs reference. When in doubt,replace. With amp ON but standby OFF check bias voltage at each pin 5 and adjust if needed. Most likely,at this point,there will be quite a severe difference in voltage with standby off vs on IF the issue at hand still persists. Adjust bias voltage with standby OFF to approx minus 32-33 volts DC.

In essence what´s described above is all there is to it really. B+ should now show 340VDC+...if not we have a bigger issue. In that case yank the powertubes out of there AND with capacitors of ample voltage rating in place start the amp up SANS powertubes and recheck B+. This should now show around the 400VDC mark approx (hence the need to replace the capacitors). If it does we have an issue with the powertubes OR their sockets. In such a scenario install a fresh set of EL-34´s AND...be prepared to set bias for these (hence that trimpot for the bias circuit).
If and when that voltage number shows up on par with what the schematic calls for...then and ONLY then put that 10k drop resistor of the rail back in place to check integrity of entire amp out.

On output transformer options then.
Yes. For an amp that carries any type of negative feedback fed off the OT secondary there MUST be some sort of ground reference involved. If not the amp has no way in hell to determin where the signal is "at" so to say.
We CAN run an amp sans ground reference if NFB isn´t in the picture,but that would still entail entire loss of voltage reference for the OT secondary,and there´s other reasons for why that is bad.
Now. For an amp that lacks NFB coming off that OT secondary we can basically hook any part of that winding to ground,as it´s AC voltage we´re talking. However Trobbins is,again,correct in that we MIGHT have an issue with NFB if that voltage problem of yours exists still as we hook that 10k resistor of the voltage rail up again.
If so first up,again with amp off,check to see if it might be that the OT wires heading for each powertube has been off at any time...and been mixed up. If in doubt just have them trade places and recheck. The OT wires are installed correct when there is NO appreciable sound coming out of it in idle except hiss.
Imperative part here though is that you HAVE to have some sort of load connected to the OT secondary. IF this is a matter of oscillation..... that translates into sound which means that the OT will see some sort of load and this energy NEEDS to be released one way or another. The most convenient way is to hook a speaker up. In fact a speaker should be connected during ALL your tests for this very reason.
Racing
Squire
 
Posts: 749
Joined: Wed Dec 21, 2011 11:07 am

Re: Geloso G1/1040 Amp repair

Postby zjokka on Tue Nov 08, 2016 1:38 am

Racing wrote:According to the schematic standby works between bridge negative and chassis ground. First up,with amp off and drained,do a DC resistence check between chassis and bridge negative. This should show zero Ohms,or very close to it. If not,replace or delete that switch entirely.

Standby switch brings up the HV so we are talking about the BY126 diodes here, right? There is continuity to ground on the negative side of the rectifier when it's out of standby (unplugged & drained!), infinite resistance in standby. The switch seems to be ok!
Racing wrote:Second up we need to isolate the powertstage per se. Simplest way to do this is to lift one leg of that first drop resistor in the voltage rail,which is of 10k. By lifting one leg of that we to the letter "kill" the rest of the amp which is exactly what we need as we want this piece up and running in proper form again.

Will do that...
Racing wrote:Me i´d just throw some 450-500V axials at it and call it a day.

Here you are clearly very much mistaken. Ok, I learned now voltage tolerance adds up in series and this lesson cost me like +€5. But there is not such a thing like throwing in 450V axials... It's like googling your ass off or driving your ass off, then pay your blood and wait a for a week or so... Ok, I found some 47µF/450V radials for €2.15 a piece.
In turn measure for resistence between pin1&8 of the powertubes and chassis. This should read zero Ohms too. Pin 1 being g3,the surpressor grid,and pin 8 being the cathode.
Then measure each screen resistor for resistence,which should be around the 470 Ohm mark.
Trobbins is absolutely right in that we these days often up that value a bit first of all and what´s more up the wattage of those resistors as well. I´d use regular 1k/5w ceramics for this.

Zero ohms or close between powertube pins 1&8 and ground!
Screen resistor I measured earlier at 470 ohms, although one looks toasted as hell... I bought some 1k 10W ceramics (they did not have 5W) which I will install tomorrow.
Start by reflowing every applicable solderjoint for the powertubes and PSU(power supply unit) in sight. I don´t care how fresh they might look,hit ´em with the soldering gun and a fresh dab of leaded solder and be done with it already.
Reason is simple.
We still have a problem. Hence why i ask you to take measurments per above. Something is amiss,and we KNOW this from schematic and experience and we need to find out why.


Sure. An oscilloscope will help but for these basic moves it isn´t called for really.

I have a scope in the basement ;-)

Make DAMN sure the fuses involved are of the correct rating. Now..there´s two "death caps" in there as well...... They are of 2,2nF and you see them in the schem on the powertransformers primary side. Cut these out and discard.

I checked the fuses, instead of 1A fuses, there were 1.25A fuses installed. I replaced them with brand new slo-blow 1A fuses. Will replace the 15A fuse too, when I find a replacement.
I removed the death caps, reflowed solder on all points on the power transformer...
Install a 3 prong mains if not done already. Ie;phase,zero and safety ground. 3 wires all in all. Safety ground SHALL go to a point used SOLELY for this on the chassis. NOTHING else goes to that point. A regular modern mains wire for a computer or whatever along those lines will work just fine. If you are to keep the mains wire as a steady state install just cut the socket off and use the wire per se. In such a case be sure to reuse the stock mains wire "locks" so the wire can´t come out if yanked.

Done! I used the chassis ground point that was used just by the two death caps for the safety ground.

Will do the rest tomorrow, kind of late now. I always have a 4 ohm speaker connected when the amp is turned on. But also bought some 10W 8ohm resistor to use as
zjokka
Peasant
 
Posts: 13
Joined: Sun Sep 11, 2016 3:37 pm

Re: Geloso G1/1040 Amp repair

Postby Racing on Tue Nov 08, 2016 12:04 pm

Keep the good work up and you´ll get to the bottom of this no doubt.

For inspiration.. :D

Image

Image

Image

One of them really small twin 6V6 model 211-A´s. Had to replace PT as it was shot and did so with a MM i had laying around for a DLR. This small piece here puts out a whopping 15VAC into 8 Ohms. U do the math :mrgreen:
Finishing touches to do,but it´s a twin mode jobbie controlled by H11 optos and what not.

Image

Image

Next up is this really Mr Handyman fumbled model G.30 from the 1940´s. Twin 6L6G powered amp. Absolutely love them old "coke bottle" tubes. This one tho,as you can see,will take a little bit more work... :mrgreen:
Racing
Squire
 
Posts: 749
Joined: Wed Dec 21, 2011 11:07 am

Re: Geloso G1/1040 Amp repair

Postby Racing on Tue Nov 15, 2016 1:06 am

https://www.youtube.com/watch?v=rfTJ18myNgg

As it´s there. Here´s a basic,practical,take on explaining the James network and how versatile it really is. Mind you this is a James with just two added switches. In essence this can be pushed to the extreme leaving you with an amp...out of this world as far as width.
Within tho this one also carries a trimpot instead of the regular "separation resistor" of the James rendering that you can with ease tailor the amp to whatever guitar,whatever guitarist and whatever surroundings.
Racing
Squire
 
Posts: 749
Joined: Wed Dec 21, 2011 11:07 am

Previous

Return to Miscellaneous British Amplifiers

Who is online

Users browsing this forum: No registered users and 2 guests