25L15 New built

[kingplank]

New built form new and recycle parts, with generous help from forum members.
At start up, smoke billowing from a couple of resistors. Turn out to be test leads wires crossed. Still very low sound. After using a capacitor signal tracer, I was able to correct a number of wiring mistakes and now it is working beatifully.

Checked all critical voltages, and I am quite happy at what they are at right now.

- I was expecting that this circuit to be extremely sensitive to the design values but swapping the values of some resistors hardly change the voltages nor the sound.
- Still trying to sort out the first T filter(mids), not much effect compared to bypass switch.
- amp is extremely low noise, zero hum. (thanks to the 2 cokes and many filters)
- extremely clean tone even with vintage test speaker on fixed bias mode. Slightly darker tone with cathode mode. I had cathode bias rather conservatively for now, 19watts dissipations vs 24watts max rating for 6L6 359v at the screens. Will crank it up later.
- temperature ckecked all the components, all at 27-29deg centigrade, except for the bisa and dropping resistors.
_talk about dropping resistors, I have 300 ohms to drop to voltage down. I am considering getting a bigger first choke and change or remove the first filter cap to drop voltage. Comments anyone?

After everying sorted out than I will get a JBL d130 to complete the project. Oh yah, cabinets too.

[HotBluePlates]

- Still trying to sort out the first T filter(mids), not much effect compared to bypass switch. ...

Copy the T-filter shown in the Bob Crooks preamp schematic.

The T-filter cuts midrange. I don't know what you're using for a bypass, but I'd think a wire from the 1st stage coupling cap to the input of the volume control would pretty much bypass the T-filter.

You'd probably have to post a schematic of what you built if you are seeking circuit-specific suggestions.

- talk about dropping resistors, I have 300 ohms to drop to voltage down. I am considering getting a bigger first choke and change or remove the first filter cap to drop voltage. Comments anyone?

... 359v at the screens. ...

What was your supply voltage without the 300Ω dropping resistor?

Again, a schematic would probably be helpful. The original 25L15 power supply is not laid out like a Fender or Marshall supply.

[kingplank]

I am glad the HBP asked me to measure the recto voltage, it is 387v which is a huge surprise for me. The same transformer running a 57 Low power Twin had 485v there and 465v at the 6L6 screens.
I have posted the as built schematic , voltages noted as example here:

TP24  = Test point #24
102v  = cathode bias mode
100v = fixed bias mode.

I have revised the T filter a bit, it works. I suspect c3 300p should be 100p
T filter bypass works

OT and second choke remained stone cold after 2 hours of playing
Power transformer , warm
First choke, almost too warm to touch.
300 ohm Dropping resistor. bbq hot, drops 37 volts. Needs some thought here. I am gun shy to remove it from where it is; but how about putting it in series with C32, the first filter cap?
The Standel schematic calls for preamp power supply bypassing the second choke L2 with a 470om resistor. Will that affects filtering or hum at the preamp?
Still amazing sound, with the available controls, a lot of variations are possible. The test is limited by a 12" Jensen alnico.

[sluckey]

300 ohm Dropping resistor. bbq hot, drops 37 volts. Needs some thought here. I am gun shy to remove it from where it is; but how about putting it in series with C32, the first filter cap?

37 volts across a 300Ω resistor is dissipating only 4.6 watts. Your schematic says you're using a 30 watt resistor, so it's fine. I would expect that resistor to be hot, but not bbq hot. Don't put it in series with C32.

I don't quite understand your cathode bias circuit, in particular, the diode across R32. And TP3 shows 21.4 volts. That voltage is across R28, a 15.6Ω resistor. That means there would be 1.37 AMPS flowing thru that resistor and also flowing thru the output tubes! That's not realistic. There must be something amiss in that simple cathode bias circuit.

[kingplank]

Sluckey you are right something is amiss. The 15.6ohn resister R28 is not there, sorry, drawing mistake.
So at 21.4v bias it would give power dissipation of 16.67v (max is 24V), I can crank this up a lot more.

Still wondering if I need the dropping resistor at all, the 6L6's are only seeing 325V

[sluckey]

So at 21.4v bias it would give power dissipation...

So where are you measuring 21.4v? The left side of R32 would be the logical place to measure it. But now the diode comes into play! It's connected parallel to R32 in such a way that there can never be more than 0.7v across R32. Is the diode really in the circuit? If so, surely it's drawn reverse polarity???

Still wondering if I need the dropping resistor at all, the 6L6's are only seeing 325V

I'd put a jumper wire across that resistor and see what it sounds like. Check some voltages too. The 6L6s can certainly handle much higher voltages. I suppose there could be some reason to run the tubes at such low voltage. Maybe HBP can comment on that.

[kingplank]

Slucky you caught me again, I did realise I've drawn the diode backwards and the 21.8 meaurement is taken on the + side of C25.
Got the diode idea from the Fender Prosonic.
I will jumper the 300ohm dropper tonight and post the results. I know that's no problem with fixed bias.

[HotBluePlates]

... The original 25L15 power supply is not laid out like a Fender or Marshall supply.

You've got the power supply mostly laid out Fender-style. You should copy the napkin schematic you posted, at least regarding that, as you have plate and screen voltages that are similar.

For reference, I have ~415v on the plates and ~315v on the screens of my output tubes. I run them around 60mA each at idle, and my PT is rated for 200mA d.c.

Then again, if you build it as per the original 25L15, you cannot run ultralinear mode because the tubes will melt in UL (because the screen voltage jumps up 100v).

I am glad the HBP asked me to measure the recto voltage, it is 387v which is a huge surprise for me. The same transformer running a 57 Low power Twin had 485v there and 465v at the 6L6 screens. ...
300 ohm Dropping resistor. bbq hot, drops 37 volts. ...

Well, the 25L15 draws a lot more current than the 2x 6L6 Twin. It's not a class A amp per se, but is a very hot class AB.

Regardless, if you remove the 300Ω without rearranging the screen supply per the napkin schematic (and/or use UL mode), you won't be able to cathode bias the amp effectively. That is, if you truly were getting 485v with the same transformer. That's because plate voltage will be too high to idle close to max dissipation, and cathode bias doesn't work great with lean AB amps.

... The Standel schematic calls for preamp power supply bypassing the second choke L2 with a 470om resistor. ...

No it doesn't. Or at least, there was no such thing in the original amp.

I read that statement as "470Ω from one choke lead to the other" which would defeat the choke being in the circuit. And if it was from either choke lead to ground, it would be a low-quality short-circuit which would also be not-good.

... the 21.8 meaurement is taken on the + side of C25. ...

21.8v/100Ω = 218mA (!) Are you sure this is right?

FWIW, If you choose not to adhere to the napkin schematic or my post of the 25L15 power section with regards to biasing, you should note that the 807's have: cathodes -> 68Ω -> 2x 100Ω in parallel -> 100Ω pot with wiper to ground. Assuming the wiper is centered, that's equivalent to a 168Ω cathode resistor.

So you might need to bump up your cathode resistor above 100Ω, especially if you have ~100mA per tube now (109mA based on math above), and are contemplating removing the 300Ω resistor which will raise plate voltage even more (and plate current along with it). In fact, since you'll probably  have screen voltage near plate voltage due to power supply layout and UL mode switch, plate current will go up even more as screen voltage rises.

Which ought to point out the danger of getting too clever with additions when trying to replicate an old circuit. My own 25L15 put me in my place in that regard when I got too clever and tried to add a Pentode-Ultralinear-Triode switch without observing that plate and screen voltages aren't the same in the original.

I also strongly recommend wiring the T-filter just like the schematics you've seen at least to start. It has a major impact on the amp's sound, and your circuit is so far removed from the vintage or modern Standel I couldn't guess what it sounds like.

[kingplank]

After a few hiccups and wiring errors, I spent the last 2 weekends testing.The amp is not blasting loud, just slightly over the Fender Blues Junior(15 watts).

1) T- filter; I put a rotary switch and have 6 drastically different caps, the first one , .001uF as used on the napkin schematic gives the most "Standel" tone. I really have fun with this switch as it makes such a variaety of tones. HBP may not like this.

2) The preamp section resisters values are not critical at all. Tried different values because I thought they were "wrong" but switch back later.

3) The amp sounded a little cleaner with fixed bias but the difference is very small. As a side note, Standel makes a version of this amp 100L15 now with fixed bias

4) The Treble bass controls work well, very un-Fender like. Tried different caps and the one that works best is on my attached schematic

5) Power supply: here is a big one here. I had originally wired it Fender style. But the guys at the forum insisted that I go with "napkin" schematic. Than suddenly all the voltages looked good and the amp sounded even better. The only thing the runs a little warm is the 270ohm resister in series with the first filter cap, at 55deg C. No worries here.
The value of this resistor is very critical as it sets the overall supply voltages.

6) after the power supply change, I checked the temperature of all the components after 2 hours of use. The first choke and power transformer: slightly warm. Second choke and output transformer: stone cold. All others 30-50deg C

7) Unresolved issued: when I cranked the amp to max, it will distort(no surprise) but it will stay distorted even after I turn down the volume. The only way to get rid of it is to turn the bypass off and turn it back on. Seems like something had saturated in there.

8) I have tweaked the bias resistors(both mode) and the values may not suit your power transformer

Here is a pieced together sound clips of my son practicing for a concert. Pictures included.
http://youtu.be/91jHVDQqD0U

[HotBluePlates]

The amp is not blasting loud, just slightly over the Fender Blues Junior(15 watts).

I haven't had a chance to turn mine above about 4. The JBL D130 is very efficient compared to other speakers and will likely bump up the acoustic volume noticeably.

1) T- filter; I put a rotary switch and have 6 drastically different caps, the first one , .001uF as used on the napkin schematic gives the most "Standel" tone. I really have fun with this switch as it makes such a variaety of tones. HBP may not like this.

If it works for you, rock on.

Originally, you seemed to want to clone the 25L15 sound, leading to my recommendation to stick with the original circuit. But if you like what your circuit is doing, how could it be "wrong"?

4) The Treble bass controls work well, very un-Fender like.

They sit in a feedback loop, so they do work very differently. Putting both controls on 0 does not kill all treble or bass, and still should yield a usable sound.

The only thing the runs a little warm is the 270ohm resister in series with the first filter cap, at 55deg C. No worries here.

When I left Iraq, it was just under 53deg C (or 127deg F). Warmer than I ever want to experience again, but the resistor may not mind.

[kingplank]

Thanks again for all your comments and advise.
It has been quite a journey to get here; to order to get that fantastic Standel tone. Yes it would have been much easier to build another Fender clone. But this is really special.
As they say, no pain; no gain.
Cant wait to build the next project: a mini 25L15. No, lets call it 10L12.

[HotBluePlates]

Cant wait to build the next project: a mini 25L15. No, lets call it 10L12.

Then you'd need a 12" JBL... 

The "L" in the model name stood for "Lansing," the maker of the D130.

[kingplank]

Still have one un resolved issue: when I crank the 25L15  to overdirve(not intentional to get distortion); it will stay heavily distorted even if I turn the volume back down. The only way to get rid of it quickly is to turn the bypass off and on again. If not using the bypass, it will stay distorted for about 5 minutes before settling down. I have never experienced this before. Not even overdriving a Fender to 19.
Does anybody have any suggestions?

[tubenit]

The only thing I ever experienced like that was an oscillation problem creating an unpleasant distortion or total blocking.

With respect, Tubenit

[HotBluePlates]

Still have one un resolved issue: when I crank the 25L15  to overdirve(not intentional to get distortion); it will stay heavily distorted even if I turn the volume back down. The only way to get rid of it quickly is to turn the bypass off and on again. ...

What's the bypass again?

I haven't had a chance to turn mine above about 4. ...

Well, this past weekend I got a chance to turn it up with Tubenit.

Both of us were playing when I had it nearly dimed, so I can't be certain how much distortion I heard while playing (maybe Tubenit can chime in). The best I can recall, the sound got a little fuzzy at the top end of the volume control but was substantially clean (especially compared to Tubenit's blues-n-rock machines!).

Not sure what's causing your distortion. I notice looking back at your schematic again that the jumpering of the 12AX7 grids defeats the different grid reference for each "channel"; unjumpering that will cause the bass of the side connected to the T-filter to jump up noticeably.

[kingplank]

I shoulda made myself more clear. Bypass switch(Fender terminology again) is the high voltage switch right after the rectifier.

There is no distortion from the 12AX7 even when the guitar volume is at set at max. (and the Standel volume control is after the 12AX7 stage.)
So the distortion at very high volume must be either the tone, inverter or power section.
Also of note: the amp stays clean right up to the point of distortion.

[PRR]

Hook voltmeter to various points, play to make it sick, see what voltage changes and takes time to come back right again.

Suspects may be power tubes' G1, G2, then try plates of little tubes.

[kingplank]

I am not able to get the to replicate the constant distortion; at least not in the stock cathode bias mode. The voltages on all the plates stayed within 5-7V even when hit hardest.
By the way, the amp sounded pretty good wide open, it sizzles with my custom made Strat with hotter than stock pickups. Still not a very loud amp. A Blues Junior can easily drowns it out.
Did some farthur experiment; with two 8ohm Celestions in parallel, 4ohm tap on the output; the amp was not happy at all. Sounded thin and constricted, and very low volume. Dont think I will do this again.

[tubenit]

the amp sounded pretty good wide open, it sizzles with my custom made Strat with hotter than stock pickups. Still not a very loud amp. A Blues Junior can easily drowns it out.

HotBluePlates amp was waaaayyyy louder than a Blues Jr. It sounded much more like a 40 or 50 watt amp.

Just an FYI.   With respect, Tubenit

[HotBluePlates]

I am not able to get the to replicate the constant distortion; at least not in the stock cathode bias mode. The voltages on all the plates stayed within 5-7V even when hit hardest.

Then it sounds like the problem is in fixed-bias mode, possibly an issue with the fixed bias supply.

Thought experiment: A large input signal drives the output tube grid from its negative voltage at idle towards 0v during the positive peak. The output tubes may be distorting mildly at this point, or may be largely clean. If the input signal is made bigger still, the grid is driven positive. The grid is always assumed to be an infinite impedance, but this is not true when the grid is more positive than the cathode; it starts drawing grid current and thereby becomes a much smaller impedance than it was when the grid was negative of the cathode. There can only be 2 sources for this grid current: the phase inverter's plate output or the fixed bias supply. The phase inverter of a guitar amp is not designed to source any appreciable current (because a design assumption for class AB1 is that no grid current is ever drawn), so the output tube grid likely sucks its current from the bias supply filter cap. That current draw drains the cap and reduces the bias, keeping the tube in overload unless the bias supply is a very much lower impedance than the grid when it is sucking grid current. A large bias supply internal impedance means it takes longer for the bias supply to recharge to its full value of bias voltage after being drained.

Okay, if that scenario is correct, I'd expect you could put your amp in fixed bias mode, attach a dc voltmeter from an output tube grid to ground, and observe the negative voltage dropping perhaps significantly. I don't know if an analog meter would be preferable to show a quick change that wouldn't be obvious on a digital readout.

Regardless, this is one of the two places PRR suggested you probe: G1 (pin 5, the control grid) and G2 (pin 4, the screen grid). If either of these are changing drastically at the moment of distortion, you have an indication of where the problem is.

[kingplank]

HBP's thought experiment makes sense, I am sure that happened.

I bought a $5 Yamaha 40W oragan speaker from my neighbour(still no $$$ for D130) and the amp souned better than ever.

We had a jam session with the 25L15 yesterday, with the guitar and amp constant maxed out at 10 (on fixed bias mode), T- filter at "stock" setup, the amp stayed clean. Voltage at pin 5 stayed rock solid at 27.4v, and pin 4 drops about 8v when I hit it hard with power chords. Still no fuzz type ditortion.

With an overdrive pedal kicked in, the amp is considerbly louder, there is distortion from the pedal of course but the amp stayed clean. Voltages obsevations more or less same as above.

It is strange that I am not able to replicate that perpetual distortion as happened 2 weeks ago. I am not too worried about it.
If it aint broke don't fix it

I will continue to post results of experiments with this amp to share wih you guys

[kingplank]

I've been using this amp now daily for 4 months. Finnally all the wiring mistakes are now fixed.
One thing that I havent tried is throwing the switch to Ultra linear. If I staying with fixed bias, is there any harm?

[HotBluePlates]

One thing that I havent tried is throwing the switch to Ultra linear. If I staying with fixed bias, is there any harm?

Your power supply doesn't 100% match what I have for a 25L15, but it does have a fairly large resistor dropping screen voltage.

When you flip to UL mode, the screen voltage will rise to equal or exceed the plate voltage (this is normal); when that happens, plate current will rise, maybe a lot. The amount of plate current rise will depend on how much lower screen voltage is than plate voltage presently.

That will mean your bias setting for normal-mode fixed-bias will likely be too hot when running in UL mode. I originally had a switch to run pentode/ultralinear/triode mode in my 25L15 but abandoned it because of this issue. For me it was more important to opeate as much like the original than to work out the issues with making it a swiss army knife.