Switching Plate Load

[Colonel_Sanders]

Hello,

I'm a newbie doing my first design and have a question. If I was going to add a parallel plate load switch with a DPDT relay to the 5F6A bassman preamp (12AU7 + 12AX7), what voltage relay would I require? I think the B+ is about 325VDC and the voltage drop across Ra is about 175V. Do I need a relay that will cover the B+ plus spikes, or is there something magical I don't understand?

Thanks,
C_S

[tbryanh]

I assume you are using a DPDT switch because you plan to switch between two resistors. In other words, with the switch in one position, R1 will have both leads in the ciruit, and R2 will will have both leads out of the circuit. Vice versa for the other position of the switch.

You might consider leaving one resistor in the circuit at all times and use a SPST switch to add another resistor in parrallel when needed. I don't know if this would help with the voltage rating of the switch.

Switches I believe are rated at a specific maximum voltage they can interupt with a specific current flowing through the contacts. (I know contactors are rated this way.) This can be an AC or a DC rating, or it might have both types of ratings. Generally, AC ratings are meant fo 60 cycles. Not sure if higher frequencies matters.

There is also the issue of inductive and non-inductive loads, and I think switches can be rated for this too. (I know contactors are rated this way.)

Since your dealing with preamp tubes, the amount of current is probably not that much of an issue, and you don't have an inductive load, so you got that on your side too.

I bet a switch for the typical standby circuit of a power amp would work fine.

[Enzo]

Yes, tbr, the one permanent resistor with a parallel method will greatly reduce the voltage needs of the contacts. You would never be interrupting the voltage or current.

In fact a SPST will be all you need. Leave the parallel resistor connected at one end. Now switch the other end in or out.

[tbryanh]

Yeah, it would reduce it from the full B+ to just what is accross the plate resistor. Good point.

[Colonel_Sanders]

Hey guys, thanks for the replies.

I am planning to have one resistor in the DC path at all times (thanks to Kevin O'Connor!). The DPDT switch is to switch a parrallel Ra in on both input stages of the 5F6a design. Would it be better for me to use 2 SPST relays driven from the same control voltage?

Thats great news on the current and voltage requirements! I'm still a little new at this, but does that mean that I'll only need a relay with a maximum rating of 175VDC/90VAC (V drop across Ra)?

Cheers,
C_S

[Enzo]

Sorry, I thought you were using DPDT to use a switch at both ends of the part.

If you are switching two separate circuits at once, my only remaining concern would be the two plate circuits would then be in very close proximity - within the switch - but I don't think that is an issue with the plate loads. After all they are right next to each other on the eyelet boards.

[Colonel_Sanders]

No worries man. Thanks for the tip, I'll be extra extra careful with the soldering.

And the relay will only need to be half the voltage of the Ra V drop because the load is in parrallel huh?

Thanks again,
C_S

[tbryanh]

Actually if the the guitar is plugged in and the amp is cranked up, the voltage accross the plate resistor can be the full B+ if your unlucky and the contacts open when the tube is saturated.

If you first mute the amp and then switch, it would always be 1/2 B+.

[Enzo]

Do you really find typical 12AX7 preamp circuits with 300v p-p across the load resistor? I don't see nearly that much signal.

Switch and relay contacts are voltage rated more for what they can interrupt than for what flows through them. SO unless you are switching while cranking a max signal through the amp, the signal amplitude won't matter.

[Colonel_Sanders]

Enzo, thanks for the info. Makes the ratings heaps clearer.

I am planning this switch, and a myriad of others in the amp, for live use, so its entirely possible that the parrallel Ra will get switched in while the tube is in saturation. That means I WILL need a relay to potentially stop all the B+? Eep, confusing.

If anybody can recommend some reading material on the subject that'd be great as well.

Cheers,
C_S

[tbryanh]

Getting the relay might be a little tough.

Relays are used alot in guitar amps, but you see them mostly in the lower voltage areas around the cathodes and tone stacks, etc.

I am sure there is something out there. Please post what you find. You've got me a little curious now.

[Enzo]

What stage in the amp is this going to be used in? You can easily measure the amount of signal on the plate of such stage. The tube only has so much gain, so if it is the input stage, I might expect as much as 30v signal rather than 300.

[Colonel_Sanders]

Just the input stage, but I'll be using a 12AX7 instead. Attached a snippet of my mock up schematic.

The amp has 3 options, but I'll just run with 2 for now.

So, for V1a when 3a is closed, the AC30/5F6A JFET's are on, making:

Ra=104K
Rk=1.5K
Ck=25uF

V1a When 3a is open:

Ra=220K
Rk=1.5K
Ck=25uF

I'd be really interested in getting the formula to calculate the stage gains and p-p voltage swings.

Cheers,
C_S

[Enzo]

I don't think it really makes much difference, but you are aware that the JFETs do not go to zero when ON? SInce they are ON until turned OFF, you can hold one in your hand and measure the ON resistance - I'd expect maybe a couple hundred ohms.

[loudthud]

There are JFETs that are suitable for switching. The J105 comes to mind. The on resistance is something below 3 ohms. It takes about 9V to turn them off and the breakdown voltage is only 25V. Should work well in cathode circuits.

[tbryanh]

You can dig up formulas for stage gain, but in a 12AX7, the most it can be is 100. Typically it is probably about 50.

The hottest guitar pickup probably puts out about a 1/10 of a volt. I don't know if this is peak or RMS. To be on the safe side, lets use 2/10 of a volt.

2/10 X 50 = 10 v

If you use a pedal that boosts the signal to overdrive the amp, and if the pedal has a gain of 10:

10 X 10 = 100v

The pedal might have a gain of more than 10.

It is always safe to err on the side of overestimating than underestimating.

It might be best to make sure the switching device can handle the full B+, and to have some breathing room, you might add an additional 20% - 50% on top of that.

[tbryanh]

If you had an inductive load, say in the output stage, the flyback voltage can be an issue, and you might get double or tripple the B+.

This should not be an issue in the preamp stages.

[Colonel_Sanders]

Tbryanh, so even though the switch is in parrallel with the permanant plate load, there is a chance it the full B+ might develop across the 2nd plate load resistor?

I do realise the JFET On/Off situation, just get a bit confused at times :-). Thanks for the 105 recomendation, gonna check em' out now.

Enzo, I'm not sure if it will make much of a difference either, but I plan to overkill first, then whittle the design down once I have it going.

Thanks heaps for all the replies!
C_S

[d95err]

Enzo, thanks for the info. Makes the ratings heaps clearer.

I am planning this switch, and a myriad of others in the amp, for live use, so its entirely possible that the parrallel Ra will get switched in while the tube is in saturation. That means I WILL need a relay to potentially stop all the B+? Eep, confusing.


If you use a switch like this, there will be a huge BANG when you flip the switch, because you are creating a sudden DC shift. Never ever switch (high) DC voltages in the signal path. To be safe, you should ensure the amp is in Standby before flipping such a switch.

I wouldn't bother with this kind of switch. If you want to change the gain or output level of the stage there are lots of other options (e.g. switching the cathode bypass cap in/out).


EDIT: Since you say you're a newbie - this design looks way too complex. It seem like you want to make an amp that can do absolutely everything. My advice is:

1) Stick to the "KISS" principle (Keep It Simple Stupid!). Build an amp that can do one thing really well, rather than one that can do lots of things badly (or not at all).

2) Stick to a well known amp design (e.g. a "Plexi").

3) Once you get one simple amp working - experiment and mod it to death!

4) Once you've modded everything you possibly can in that amp - start designing your dream amp

(Hope you don't take this the wrong way, this is just some friendly advice based on my own experiences)

[Colonel_Sanders]

I know where you're coming from. Last year I didn't know what a capacitor was! I do want to design my own amp for my first build, but if there are simpler ways to achieve what I'm trying to do I'm all ears. Can you point me to a site or book that can explain how I can calculate the gain of a 12ax7 stage, and with and w/o the bypass cap? Not quite sure what to look for...

Thanks,
C_S

[tbryanh]

Yeah, if you use a mechanical switch, your probably going to get a big pop during switching.

If you use a jfet or mosfet, you can probably use an RC time contant to gradually make the transistion and avoid the big pop.

I haven't seen any circuits that have switches in the plate section, so I am a little curious about it.

As d95err asked, what is it your trying to accomplish?

Changing plate resistors probably changes the characteristic curve of the stage and therefore the distortion effect; i.e. different plate resistors probably provide different distortion effects.

Some amps mix it up. The first stage uses a 100K, the second stage uses a 220K, the third stage uses a 100K, etc.

I heard that mixing it up give a fuller body distortion.

[Colonel_Sanders]

I'm trying to design an amp that can mimic 3 different amp's, using the same expensive components (trannies, tubes etc.) in each design. Needs to be switchable for live use.

Seems like I'll have to go with an alternative:
What about an LDR in parrallel with the plate load?

I like the idea of a fet and ramp, but the switch would still need to be rated at 300V+, yes? Is it possible to parrallel mosfets for higher voltage/current ratings?

Thanks,
C_S

[Enzo]

WHy not take an existing amp, tack a resistor to the B+ and a switch between the other end and the plate and flip the switch back and forth. What happens? We have a lot of rationalizing going on here, but why not take five minutes and just find out?

Switching the parallel resistor in and out would not in my view create a large DC shift. Unlike a cap that would have to charge, the resistor is already at B+ potential before the switch is flipped. The only difference I see would be the small change in plate voltage due to the current through the tube. The old "IR drop."

[Colonel_Sanders]

Great idea, why didn't I think of that... I'll pop one my bassman tomorrow and post back with results.

Thanks,
C_S

[Colonel_Sanders]

Sorry about the delay there. Anyway, I put a 240VAC DPDT switch with a 820K resistor in parallel with the 100K anode load of my Bassman Re-issue. Definitely 'pops' when I hit the switch. I couldn't really hear a difference, but there 'might' have been a small decrease in gain. It was only a 100K/90K parallel difference though, not the 220K/100K I'm going for.

So, any ideas on how to tame the pop? And do you guys have any advice for getting a measurement of the max voltage swing while operating? I tried with my multimeter in AC mode but it just jumped around like crazy.

[Enzo]

Meter jumps around? Are you measuring with a steady sine wave test tone? No reading will be stable unless the signal is.

Did you wire the extra resistor to one end and switch only the remaining end? Or are you switching both ends of the resistor in and out? One end should be soldered on, ther is no reason to switch both ends.

If you want 100k/220k, there is no resistor you can parallel with a 100k to make it larger. STart with the 220k, and switch something parallel to it to lower the total to 100k. Just a second 220k - a standard value - would result in 110k total, which ought to be close enough.

[Colonel_Sanders]

I am switching both ends, good point. Also, I didn't think of the steady tone, was using my ipod lol. I'm going to try a 1kHz tone out of my laptop and rewire the switch now. Cheers for the tips.

The 220K will be the stock load in the amp I'm building (like the schematic).

C_S

[Enzo]

The one end has a better chance of less pop, but I am interested in your results. I have never switched a plate load resistor. But in switching cathode bypass caps in and out, maintaining a DC charge on the cap is important.

it may be difficult to eliminate the current shift and resultant noise. One thing comes to mind would be using something like a Vactrol instead of an actual switch.

[Colonel_Sanders]

Its still gives a good pop when I switch it in. I've got the B+ end connected and the anode being switched. Max voltage swing I could get from my laptop was 17VAC measured from the anode of plate load to the anode of the tube. Is this the correct place to measure?

If I can't get rid of the pop I guess I'll look into Vactrols. Are they pretty easy to control?

BTW, my filter caps drain themselves. Is that normal? (still short the caps for safety though)

Thanks again,
C_S

[Enzo]

I don't know that a Vactrol would handle the voltage, you'd have to check the data sheets. If they would be suited, they are easy to control, they are used for channel switching in lots of amps. I just meant it conceptually - as in a solid state relay sort of thing.

The DC voltage across your switch should be of course zero with the switch closed, and whatever is across the permanent plate resistor when open.

Signal level would be with respect to ground, but I am not sure what you are up to at the moment.

[Colonel_Sanders]

I'm not sure either :D. I think I'm trying to calculate the maximum voltage swing across the plate load.

According to my multimeter, my laptop is putting out 270mVAC. In the first gain stage I measure a drop of 102VDC across the 100K plate load. From the plate load to ground I measure about 17VAC pk to pk which gives the stage a gain of around 60.

B+ for V1a is 370VDC. That means the maximum voltage swing will be from 285VDC to 251VDC. As tbyranh said earlier, a boost pedal might have a gain of 10 = 2.7VAC input which means the signal swing could be about
370VDC to 106VDC, worst case scenario.

So the switch/vactrol would have to rated for 264VDC.

Does this sound about right?

[Colonel_Sanders]

BTW, there are a couple of Vactrols with a cell voltage of 300V and one with a 500V rating.

http://www.ortodoxism.ro/datasheets/perkinelmer/VT500.pdf

Cheers,
C_S

[Colonel_Sanders]

Well I just bought a couple of VTL5C6's (250V cell voltage) to try out. On resistance is 2K @ 40mA and off R is 10M. I was thinking about using it as the parrallel resistor, but I reckon I'll just use it as a switch.

Thanks heaps for all your help guys. This forum is stonkin'.

Cheers,
C_S

[Gregg]

So, any ideas on how to tame the pop?

The only way to avoid poping is to use some sort of "smart" type of switching. You can find such switching in Peavey amps - the so called clamping where a triac is used.
In my amps/preamps all relays are connected to a PIC12F629 /or 16F690 if there's more to switch/ which via relay shorts the sound to ground for 50-60ms while the relays are switching thus avoiding any poping.

[Enzo]

That mutes the amp briefly when channel switching, but many guys object to the "delay" it causes. That momentary blank in the sound bugs them.

[Gregg]

That mutes the amp briefly when channel switching, but many guys object to the "delay" it causes. That momentary blank in the sound bugs them.


That's entirely their problem. I don't object even a 100ms delay if completely noiseless switching is desired. Unfortunately in such cases there's always a compromise to be made.