Looks like the article I posted is lifted from that article you posted, but edited.
Enjoy the benefits of a premium membership!
Upgrade Account
-
TalkBass has been independent since 1998. Add your voice.
Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
Join free Log in
Want zero display ads or expanded classifieds tools? Compare plans.
Demystifying Tube Amps---The Standby Switch
- Thread starter JimmyM
- Start date
- Feb 12, 2006
- 61,031
- 5
- 220,916
- Disclosures
- Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
If an amp is specified to operate at nominal 117V or 118 but can not handle 120V and the D.C. Voltage exceeds the cap voltage, that's a design problem not a line voltage problem. Nominal 117V +10% is 128V.Some of the old British amps have problems on US main voltage. Early 60s Vox AC30s were tapped for 115V, but a lot of people recommend running them on 110V. I have a 62 or 63 and it biases perfectly at 90% on 110V with an 82 ohm cathode resistor. The original cathode resistor was 50 ohms. The DC operating voltages are still higher than the schematic lists. Heater voltage is 6.25V, which is slightly low but within spec. I run this amp on a variac, but the rest of my collection is setup for 120V. Some of them sound better at 115 though.
My 76 Matamp GT120 PT also runs really agro on US mains although the tap is labeled 118V. I bought it in Germany and it was still running all the original filter caps. It started having problems after a couple years here in the US. When I opened it up, I noticed many of the caps were erupting. I checked the voltages and they were over the cap ratings. Plate voltage at idle 516V on 120V line. I did a full cap job, upscaled some of the caps to a higher voltage rating, and also increased the size of the resistor feed power to the preamp voltage ladder from 10K to 16K. I believe old Orange amps have the same problem as Matamp on US mains.
I also have a couple of Hiwatts. I haven't personally worked on the DR103, but I did check the bias on the DR201. The PT tap is labeled 117V and the plate voltage is 711V at 120V line, which I believe is way above spec. The tubes are biase a little cold though and there is no adjustment.
I have a 70s 100W JMP and an 80s 100W JCM800, and they are both fine on 120V 60hz.
I agree. I think whoever did the calculations for the power transformers did not get things quite right. The theory I have seen in a couple of places is the turns ratio was calculated for correctly for 50hz operation, but the voltage runs a bit higher on 60hz. No idea if it's true. If it is it represents a pretty silly mistake.
A power transformer have a free running voltage and a nominal voltage. When you put a transformer into nominal load it was designed for the voltage without the load will be will be anything between 10-25% more depends by its quality. With other words the voltage under load will drop 10%(best OT) to 25%( very mediocre one). So supposed we need a supply of 450v at 400mA and have a tranny which give us 450v at 400mA which have looses 20%.we can expect to have 540v when power on (standby off) , 500v at iddle , and less than 440 when full throtle ( cause for a load which draw 400mA DC it need a tranny with 600mA AC capabilities if use a bridge rectifier. Add to this picture a poor mains regularisation and will found a good design should use a 600v rated cap at least
Any transformer specs are described by nominal voltage at the max AC current was designed. A random quality PT have looses around 15%.
Last edited:
Just checked my GT120 notes. The two highest voltages on the preamp power supply ladder were 476V and 455V. The original capacitors were rated for 450V and I replaced them with 500V. I also increased the value of the 1st preamp power supply dropping resistor from 10K to 16K, which brought the 476V reading down to 450V.
I found a thread where someone had taken measurements on a GT120, and the respective voltages for the same test points were 427V and 411V on EU voltage.
An important factor is the GT120 does not have a standby, so the voltages will run even higher until the tubes start conducting. Hopefully I have built enough headroom in with the component changes I made.
I could increase the resistor value more. The orange OR120 is very similar to the GT120 and uses a 33K resistor to feed the preamp supply ladder. This drops the voltage to 370V.
On a good schematic you will found both dc voltages mentioned at checking points (meant v iddling and v nominal power) . Many schematics mention only one without to mention in what conditions those readings was picked up. Worst ,found the values in some schematics are mixed.
A good example is schematic of B15 Heritage thinking you.re more familiar. Take a look how a good print job was done.
This is also a reason to put and use a stby sw. At least when is comuted "on" the voltage on supply rail will drop in respect of how much the tubes draw at iddle condition. This will be into AC side, before the main cap, for solid state rectifier. For a tube rectifier I think a stb sw it is absolutely unnecessary, but this is just my thought
Last edited:
Well, you are the authority on amps! Thing is, my wall volts tend to vary and usually above 120 not below...as well, slightly lower wall volts tends to tame the sharp top end on some of my amps.Which is very much appreciated by the sound guys
Few tube amps have a relay, though there are a few that have a time delay warm-up before the B+ (HT) is applied to the plates/screens. At one time, I had designed an amp that allowed the B+ to ramp up after 15 seconds, which was elegant but overkill. It did eliminate issues with high voltage being applied as an impulse to the plate circuit.
The savings is probably about 50% when in standby. For a larger amp like an SVT, that's significant.
Some power is dissipated in the heaters, but about 2/3 is dissipated in the plate circuit (talking about the power amp portion of the amp). On a 6550 (per tube basis), the heater power is about 10 watts, the quiescent dissipation in the plate circuit is about 20+ watts, so for 6 tubes you will save 20 x 6 - 100 watts. The hidden savings is in the reduced operating temperature of the output transformer primary and the extension of tube life due to lower operating temperature while in standby.
I don't think so... see the above calculation.
It's the time required for the cathode to heat up enough to begin boiling off electrons.
Agreed.
This is not necessarily true either. Many older amps were designed for 117V and 120V nominal operation and pose no problems whatsoever. The 1969ish schematic set I have shows the Ampeg SVT and B-15 were designed around 120V, The Oliver amps were 115/117V, and most of the Fender amps were 117V which later became 120V in the CBA and post-CBS eras, Musicman was all 120V, Rivera was 120V, Sound City was 115V, the old Trace V series was 120V. Even my Ampex service binder shows all amps built in 1956 were set up for 117V. None of these amps will have a problem with nominal 120V.
In fact, I'm having trouble finding ANY in the mainstream that are less than 115V.
If you go back to the 1930's and 1940's, you may find more, but there weren't many amps designed for bass back then.
There are certainly some, but in general they represent the solid minority.
Tonight, the wall volts are around 122-123 range measured with a Fieldpiece SC66 and a few other meters.
62 Bandmaster....around 495 plate volts at my wall volts which as i stated earlier varies at 118 to sometimes 125 but usually between 118 to 122...so then, if i dont bias at around 70% dissipation the plate volts go over 500....many 60's Blackface Fender Deluxe and Princeton have very high plate volts for a 6V6GT amp as well.Some of the old British amps have problems on US main voltage. Early 60s Vox AC30s were tapped for 115V, but a lot of people recommend running them on 110V. I have a 62 or 63 and it biases perfectly at 90% on 110V with an 82 ohm cathode resistor. The original cathode resistor was 50 ohms. The DC operating voltages are still higher than the schematic lists. Heater voltage is 6.25V, which is slightly low but within spec. I run this amp on a variac, but the rest of my collection is setup for 120V. Some of them sound better at 115 though.
My 76 Matamp GT120 PT also runs really agro on US mains although the tap is labeled 118V. I bought it in Germany and it was still running all the original filter caps. It started having problems after a couple years here in the US. When I opened it up, I noticed many of the caps were erupting. I checked the voltages and they were over the cap ratings. Plate voltage at idle 516V on 120V line. I did a full cap job, upscaled some of the caps to a higher voltage rating, and also increased the size of the resistor feed power to the preamp voltage ladder from 10K to 16K. I believe old Orange amps have the same problem as Matamp on US mains.
I also have a couple of Hiwatts. I haven't personally worked on the DR103, but I did check the bias on the DR201. The PT tap is labeled 117V and the plate voltage is 711V at 120V line, which I believe is way above spec
I have a 70s 100W JMP and an 80s 100W JCM800, and they are both fine on 120V 60hz.
It doesnt take much 'wall volt' increase to swing the plate volts and current draw upward either.
Thanks. See post #85. If the new 500V caps I installed start to
, I will pull the tubes and take some measurements to see what the unloaded DC voltage is. This will ensure the new caps I order have a sufficient voltage rating. This amp doesn't have a standby and the plate voltage is running 516V at idle, so the 500V preamp caps are probably being exposed to over voltage for a short period every time I turn the amp on. However, the original 450 volt caps should have been exposed to the same over voltage problem on EU mains, and they were fine for over 40 years, until the amp was brought to the US where the 450V rating was exceeded continuously during operation.
I have a vintage schematic for the Matamp that provide two sets of voltages, but most of the values are too blurry to read. Rather than idle and nominal power, I believe the voltages are turn on and idle. The HT voltages are listed as 516V and 500V. My amp idles with the plates at 516V, but I know the voltages in the amp are running high due the over voltage problem in the preamp voltage ladder. However, it's not unusual to find GT120s running even higher plate voltage based on various threads I have read. FYI the schematic show 180V for the HT windings, which I believe calculates to about 505V.
There are no shunt resistors in the preamp voltage ladder to pull current until the tubes start conducting, so the highest voltage will be be based on the time constant for the capacitors to charge. AFAIK it takes less than 30 seconds for 12AX7s to start conducting, but I believe the capacitors will be pushed above their limits for a short time.
The other schematic I have is from someone who went through his amp and documented working voltages with the amp biased and idling. This schematic list 467 volts at the center tap of the OT.
okcrum
in your chest
Right. 6146 was a small ruggedized design for mobile use, using a low temperature heater/cathode for controlled warmup. With tubes like 8974s, having directly heated filaments (no cathode), you have to ramp up the filaments over a 300 sec profile (to prevent fracturing the filament or its supports) before applying HV (20+ kv) or you strip the filament:
In some small transmitting tubes, like 3CX800A, you can even strip the plating off the control grid by applying HV too soon.
Last edited:
I've seen it suggested that Fender used Standby switches on their larger amps like the Bassman in the 50s because the filter caps had a voltage rating not much higher than the B+ voltage of the amp, so applying power before the valves were drawing current might exceed their rating. Then Marshall drew heavily on the Bassman design and everyone else followed the convention set by those two big players.
okcrum
in your chest
This is also true, for filament life, in continous operation like a public service transmitter. It's also recommended to drop the filament to 5V when it's used in power oscillator circuits. With only 700-800V max on the plates, and controlled warmup, you don't have to take special precautions at turn on.
Last edited:
I also consider is not a reason for concerns as time many Fenders runs also the caps aso for years, The ripple is minimal at iddle so the capacitors will not be stressed. Also the caps voltage is working voltage, true for certain conditions, but if you check the data sheets for modern caps will see where absolute ratings are. The elcos are consumable components in a PA amp and see how usualy there are mounted in a position to afford quick replacement as well as tubes are.Thanks. See post #85. If the new 500V caps I installed start to
This amp doesn't have a standby and the plate voltage is running 516V at idle, so the 500V preamp caps are probably being exposed to over voltage for a short period every time I turn the amp on. However, the original 450 volt caps should have been exposed to the same over voltage problem on EU mains, and they were fine for over 40 years, until the amp was brought to the US where the 450V rating was exceeded continuously during operation.
I have a vintage schematic for the Matamp that provide two sets of voltages, but most of the values are too blurry to read. Rather than idle and nominal power, I believe the voltages are turn on and idle. The HT voltages are listed as 516V and 500V. My amp idles with the plates at 516V, but I know the voltages in the amp are running high due the over voltage problem in the preamp voltage ladder. However, it's not unusual to find GT120s running even higher plate voltage based on various threads I have read. FYI the schematic show 180V for the HT windings, which I believe calculates to about 505V.
There are no shunt resistors in the preamp voltage ladder to pull current until the tubes start conducting, so the highest voltage will be be based on the time constant for the capacitors to charge. AFAIK it takes less than 30 seconds for 12AX7s to start conducting, but I believe the capacitors will be pushed above their limits for a short time.
The other schematic I have is from someone who went through his amp and documented working voltages with the amp biased and idling. This schematic list 467 volts at the center tap of the OT.
Yes I realize it's a bit of a dance between the plate voltage and how much current the tubes are pulling. If the voltage is high regardless, I think it's a good idea to run the tubes a bit on the cool side, especially if you have a nice stiff B+ supply, so varying the current makes little difference.
The problem with some amps is the single rail power supply, so the screens are only a few volts below the plates. At least my Hiwatt DR201, which has 711V on the plates, has a dual rail power supply so the screen are run at about 250V below the plates.
I didn't set up this amp and it does not have variable bias. I did check the cathode current and the hottest tube was only pulling 22mA, so it is set up pretty cold. The highs are pretty hard and brittle, so it might be interesting to install a variable bias control and play with it. But at such a high plate voltage, it might be risky to increase the current further, which is probably what is needed to soften the response.
I do plan to take a wait and see approach, and I am optimistic...but my intuition is not always correct
.
okcrum
in your chest
I do plan to take a wait and see approach, and I am optimistic...but my intuition is not always correct
I've found that the power transformers in older gear are the second weakest link, since they used varnish for wire insulation before more modern materials like formvar. After enough time, heat, and power cycles, the varnish dries out and cracks, and you get shorted windings.
From my experience, not the voltage kill the screens even for the worst modern tubes, the screen power dissipation will! Be absolute sure the screen resistor will limit the voltage when the amp is dimed at absolute max 10w in respect with screen current draw.( I set the max dissipation at 12w in my Marshall, but we are nuts...). You can keep the screens at 500v and you.ll be cool for any tube from modern production. Just be sure the screen resistor to be big enough to limit the dissipation no more than 10w when full throtle.
Don't forget there are also under mechanical stress due to cycle vibration.
okcrum
in your chest
Running the screen voltage as high as possible gets you also the highest possible power gain from the output tubes. Some amp designers will trade tube life for power by keeping the average screen and plate dissipation within limits, and allowing it to exceed them on peaks like note attack.
My reasoning is the hotter you bias the tubes, the closer the screens are to their limits before you even play a note. I think almost all of my amps are running stock screen resistor values, and I rarely push any of them hard enough to have any concern.
I usually set the cathode current around 70% of the plate rating, so it's a fairly conservative setup. If the amp has specific directions that are more conservative, like the Ampeg SVT or Orange AD200B for example, I will follow that. If the amp sounds like crap, I'll play with the bias a bit. But I have a fairly large collection, so I tend to be pretty conservative.
I checked the surge voltage limits of the original CDE 450V caps I put in the amp and they are surge rated to 500V...these were $2.03 each. There is still one 450V cap at the very end of the voltage ladder. I can't find the transaction for the two 500V caps. I think I bought Vishay Spraque and these run $18 a piece. The datasheet says they resist surge voltage, but a spec is not provided. Only time will tell.
For fun I did a search for 10uF 600V caps. Vishay Spraque was the only option and they were $37.14 each.
