Gates SA70 with 5840 tubes? Need help with biasing.

[rock soderstrom]

Hi guys, I can't sleep and am using the time to do a bit of planning on a new project.

I have a lot of 5840 miniature pentodes and am trying to find a useful application for them beyond microphones.

I like the Gates SA70 mic preamp and am trying to convert it to 5840 tubes. These are kind of similar to the 6j7 as a triode and the 6c5.

My planned B+ is 150V. The original is (240V).

How would you bias the tubes with anode and cathode resistors?
Can I just use the original values? I suspect that will need to be adjusted. Any suggestions?

Here is my circuit sketch based on the original.

 

[emrr]

180-280 I think is the B+ rate mentioned. 180 being RCA standard at one point. In some instances I run my originals at 365. Try the stock values.

 

[abbey road d enfer]

HHow would you bias the tubes with anode and cathode resistors

Data sheet suggests Rk=560 ohms, but it's probably for VHF operation, where optimization is different than audio frequencies.

Can I just use the original values? I suspect that will need to be adjusted.

You will have to. You need to start somewhere, so why not with the original values?

Any suggestions?

The 1st tube bias must optimize noise, so you'll probably need to run it a <1mA anode current. For the second tube, you'll need to optimize THD and output level, so probably run it at slightly higher current, but with 150V and Ra=56k, you don't have much wiggle room. You'll probably end up with 1.5-2mA

Here is my circuit sketch based on the original.

This circuit is a bit weird, since the gain control is included in the NFB loop. Since the NFB circuit is selective, its action is not constant wrt the gain setting...

 

[emrr]

1.4 and 2.3 mA with 6J7 and 6C5.

 

[abbey road d enfer]

1.4 and 2.3 mA with 6J7 and 6C5.

Of course, that's what the diagram says, but with only 150V for B+, it's not simple, and probably not so good for 5840.

 

[rock soderstrom]

Thank you guys for your feedback!

I think I'll change my low B+ voltage and do it closer to the original. The idea was to make the preamp as small as possible and use a quadrupler to make an DC voltage from an external wall wart to B+ (as well as phantom power and heating from the same AC source). I also wanted to keep the power transformer away from the signal transformers.

I'm now planning on using a small internal toroidal transformer, let's see how big that gets. I will cancel phantom power.

Yes, NFB of the Gates SA70 is unusual, but it works. I will make NFB switchable, the SA70 is on the character side anyway, see if I can get away with it. I left out the second capacitor in the NFB branch, both caps are there to "adjust" highs and lows to the input transformer. I let my ears decide how much of it I need...

Edit: Is there anything wrong with applying NFB directly from the V2 anode to the V2 grid?

 

[abbey road d enfer]

I think I'll change my low B+ voltage and do it closer to the original.

Then you may exceed specs. Datasheet says max anode voltage 165V. You don't have much legroom.

Edit: Is there anything wrong with applying NFB directly from the V2 anode to the V2 grid?

The issue there is that voltage-to-current NFB reduces the input impedance, which may alter too much the potentiometer taper.

 

[emrr]

Of course, that's what the diagram says, but with only 150V for B+, it's not simple, and probably not so good for 5840.

Nowhere does anything official say 150V, that's a typo or a change in this thread. 180-250 is what the official documents mention, with 100V roughly at each plate on the drawing example. 5840 looks fine for this, data sheet gives 100V plate as the operating example. Can't hurt to burn a pair down and see!

This is a very popular circuit. I'll claim starting the internet awareness on them over 20 years ago, the Coil CA-70 version is selling as fast as they can build them today.

 

[abbey road d enfer]

Nowhere does anything official say 150V, that's a typo or a change in this thread. 180-250 is what the official documents mention, with 100V roughly at each plate on the drawing example. 5840 looks fine for this, data sheet gives 100V plate as the operating example.

IDK.

 

[emrr]

I had an RCA BA-2 someone had re-socketed for triode EF-86's. I did some listening tests, checked the amp measurements, then converted it back to 6J7. Could hardly tell the difference, transformers dominate sonics. Tangent thought there.

Oddly, I've never liked a 6J5 in place of a 6C5 in an SA-70, though they are thought of as interchangeable with many amps specifying either type.

I got curious, as Gates doesn't really spell out specifics in their consoles as many others do, just ranges in the individual amp manuals. It looks like 250 B+ is what actually happens in the SA-40 console.
≤ 100V at the plates, 3.4mA with 240V on the schematic.
My originals run about 290V usually, more like 5mA in reality (25mA for 5 channels).

5840 heater-cathode is well within the safe zone, looks actually double the original tubes.
Plate dissipation is about half, but not a problem in this range.
Plate currents in examples are vastly higher than would be encountered.

I went old fashioned and rebuilt an SA-70 with a 36 and a 37 tube once. I did change the cathode on V1 by looking at the clipping point with varying resistance. Better? Worse? Can't say. It worked as well as the others, and I did not try the same observations on varying cathode in the originals.

 

[rock soderstrom]

This is a very popular circuit. I'll claim starting the internet awareness on them over 20 years ago, the Coil CA-70 version is selling as fast as they can build them today.

I am a repeat offender, I have already built the SA70 with original tubes twice.

No doubt, without you I would never have heard of this fat sounding amp.

Here is the only info I found regarding 5840 in triode mode.

 

[My3gger]

I think I'll change my low B+ voltage and do it closer to the original. The idea was to make the preamp as small as possible and use a quadrupler to make an DC voltage from an external wall wart to B+ (as well as phantom power and heating from the same AC source). I also wanted to keep the power transformer away from the signal transformers.

I'm now planning on using a small internal toroidal transformer, let's see how big that gets. I will cancel phantom power.

Yes, NFB of the Gates SA70 is unusual, but it works. I will make NFB switchable, the SA70 is on the character side anyway, see if I can get away with it. I left out the second capacitor in the NFB branch, both caps are there to "adjust" highs and lows to the input transformer. I let my ears decide how much of it I need...

Do you want several channels in small chasis, or one very small to use with wall wart?
Someone used 2xEF86 and ECC82 for dual channel, 6CG7 might sound better (bigger anode, etc).
Some toroid winders will wind toroid "more vertically", making it easier to put in tight space. mine can be much closer to sensitive parts than EI or some other toroids not winded for sensitive audio.

 

[abbey road d enfer]

Here is the only info I found regarding 5840 in triode mode.

Enough to deduct Ra =10k and s= 6mS

 

[rock soderstrom]

Do you want several channels in small chasis, or one very small to use with wall wart?

My first idea was to build just one channel, as small as possible with external transformer (wall wart), but now I think it much easier to include a small toroidal transformer and skip all the PSU hassle.

Some toroid winders will wind toroid "more vertically", making it easier to put in tight space. mine can be much closer to sensitive parts than EI or some other toroids not winded for sensitive audio.

Sounds very interesting. Do you have some photos of these kind of transformer? Source?

This is the transformer I will use.

 

[rock soderstrom]

Enough to deduct Ra =10k and s= 6mS

How do you do that?

 

[My3gger]

My first idea was to build just one channel, as small as possible with external transformer (wall wart), but now I think it much easier to include a small toroidal transformer and skip all the PSU hassle.

Sounds very interesting. Do you have some photos of these kind of transformer? Source?

Winder is local doing audio power trafos more as a favor meaning local pickup, discussions, etc. While ordering one for me i can ask for the tall one at 16VA and send you one, they are cheap. I though most winders do this...
This are approximate measurements for common trafos, i have more information about them:

Power​	diameter D​	height H​	weight​
VA​	mm​	mm​	kg​
20​	60​	30​	0,4​
30​	69​	32​	0,5​
​	80​	25​	0,5​
50​	75​	38​	0,7​
​	79​	33​	0,7​
​	90​	28​	0,7​
​	65​	50​	0,7​

Schematic is nice, easy for understanding NFB and filters, thanks.

 

[Heikki]

How do you do that?

Radiotron Designers Handbook third edition, chapter 34 (pdf page 273).

 

[mhelin]

For one channel you could try also tubes like ECF80x/8x (pick one with the lowest output impedance triode and the highest gain pentode) or dual triode 12DW7 (ECC832 =12AX7 + 12AU7 halves, see JJ Electronic - ECC832 - 12DW7). Couldn't you also adapt the REDD.47 gain setting/feedback into this circuit, I mean adjustable feedback from the output triode's anode into the input pentode's cathode (in general, no using the exact filter network or fine tune, just as simple as possible)?

 

[rock soderstrom]

For one channel you could try also tubes like ECF80x/8x (pick one with the lowest output impedance triode and the highest gain pentode) or dual triode 12DW7 (ECC832 =12AX7 + 12AU7 halves, see JJ Electronic - ECC832 - 12DW7).

Thanks for the good suggestion, that would certainly make sense but I have another plan.

As described before I want to use the 5840 tubes, because I have a lot (about 30) of them.

That's why I came up with the idea to build mic preamps with these tubes.

If I would not build with the 5840, I would take the original tubes 6j7/6c5, which are very good and still cheap to buy. Secret tip!

Your idea with the negative feedback from REDD47 sounds interesting, I habe to think about it.

 

[rock soderstrom]

Winder is local doing audio power trafos more as a favor meaning local pickup, discussions, etc. While ordering one for me i can ask for the tall one at 16VA and send you one, they are cheap. I though most winders do this...
This are approximate measurements for common trafos, i have more information about them:

Power​	diameter D​	height H​	weight​
VA​	mm​	mm​	kg​
20​	60​	30​	0,4​
30​	69​	32​	0,5​
​	80​	25​	0,5​
50​	75​	38​	0,7​
​	79​	33​	0,7​
​	90​	28​	0,7​
​	65​	50​	0,7​

Schematic is nice, easy for understanding NFB and filters, thanks.

Thanks My3gger for the Info and the offer! Sounds interesting. I will PM you this week.

Cheers