• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Greetings and a couple of questions

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Greetings! I have lurked on this forum for a while and enjoyed reading the high quality posts and great willingness to experiment compared to some of the other audio sites.

I look forward to posting now and again. I have a tube hobbying background of about three years, but no formal education in electronics. The bulk of my knowledge came from reading the early NEETs manuals and absorbing things from internet sites, including Steven Bench's, and much trial and error. I am a cheapskate and tend to use salvaged parts and tubes garnered from retired TV repairmen and old consoles.

I have several cases of NOS and used tubes and am about to embark on my first DHT SET project. I have a bunch of RCA and Sylvania 6b4gs, which I hope to drive with 6sn7s running at 10-15 mA. I picked the 6sn7s because I have a lot of them. I have had really good experience in the past with cathode battery bias of the first stage and grid battery bias of the output stage (I have not heard any benefit to an extra driver stage, so I have kept most amps to two-stage).

To the main topic. Cheap ($2) 12v camera batteries from WalMart soldered in series make great grid-biasing batteries. The technique is to stick a 100K resistor in series with the battery and the ground and ground the cathode. I find that this sounds superior to the bypass capacitor necessary for cathode bias and it's cheaper than rechargeables with sufficient ampacity for cathode battery bias. Energizers solder together well.

My first question involves my difficulty of thinking about what happens when the previous stage swings above the available bias (say -48V for four batteries). The battery is now going to draw a current, but the current will be miniscule because of the 100K resistor. I've done this many times in the past and never had a problem with batteries exploding. Is there something dangerous to this practice that I am not seeing?

Another question is whether anyone has tried the Edcor 15W 5K:16 transformers with 2a3/6b4g tubes. They should have an ideal load impedance of 2500 ohms when connected to 8 ohm speakers.

Thanks in advance.
 
Administrator
Joined 2004
Paid Member
Interesting idea using camera batteries for bias. I assume you are using them to provide -bias to the grids? In this configuration you would ground the grid, normally ground the + terminal of the battery and connect your grid bias resistor (100K or greater typically) to the node between your coupling cap/grid stopper resistor. If I understand you correctly in your description of battery hook up you are unnecessarily imposing ac signal on the battery which means that its capacitance to your chassis will result in a lower HF rolloff than the more conventional connection.

These batteries should not be used in the cathode circuit as I am not sure how much current they will tolerate..

In terms of your SE amplifier project, the 6B4G would typically operate in class A and not draw an appreciable amount of grid current, so your battery will not have to provide grid current in normal operation.

You might want to provide pin jacks to monitor the battery voltage from time to time and to measure the cathode current in the output tube as well.

You will need to heat the 6B4G filament with dc to get acceptable levels of hum at the output.

Also avoid the ersatz 6B4G (IIRC Sylvania) from the late 1970's/early 1980's, it's just a triode connected sweep tube with the heater and cathode also strapped together. George (Tubelab) probably knows the one I mean off the top of his head.. 6AV5?

See if you can locate a set of 3K transcendar transformers on eBay, they should be an excellent match to the 6B4G, sound great, and are quite inexpensive.

Welcome to the wonderful world of tubes... :devilr: :D
 
George (Tubelab) probably knows the one I mean off the top of his head.. 6AV5?

Yes. Sylvania stuffed some triode wired 6AV5's into some straight sided bottles and called them 6B4G's. They don't sound half bad, and can handle far more power than a 6B4 ever could, but they are not 6B4's. They aren't even DHT's. They can be easilly spotted, they have cathodes, screen grids, beam forming plates, and other things not found in a DHT.

Another question is whether anyone has tried the Edcor 15W 5K:16 transformers with 2a3/6b4g tubes. They should have an ideal load impedance of 2500 ohms when connected to 8 ohm speakers.

I did the next best thing, put 4 ohm speakers on a pair of 8 ohm Edcors. If you only have $20 each to spend on transformers, by all means do it. They are decent transformers for the money. At the power levels you are going to get from a 6B4 they won't saturate either. I tried them with a 300B, and some of the previously mentioned 6AV5's, they sound far better than $20 transformers should.

I can also agree with Kevin on the Transcendars. I bought a pair on Ebay a few years ago. I liked them enough to buy 10 more pairs. I just put a pair in a KT88 amp, and only have a very few left. I noticed he is now selling "budget SE transformers" that should beat the Edcors, but I haven't tried them yet.
 
Thanks for the reply. This is what I had in mind (minus undetermined resistor values). The 6b4gs I have are double plates. Do the fake ones have double plates also?

When you say that I am unecessarily flowing AC through the battery, is there a way to prevent that with fixed battery bias?

Thanks,

Steven
 

Attachments

  • battery bias 6b4g.jpg
    battery bias 6b4g.jpg
    24.1 KB · Views: 376
6B4 circuitry looks OK. TL rates to be right on the need for DC heating. Biplate construction sounds like genuine DHTs to me.

You don't need a fixed resistor from grid stopper to ground on the 6SN7. However, gain rates to be inadequate for full power with a CDP providing the I/P signal. Do you have 12AV7s or 5965s in your "pile"?

BTW, is that a decoupling cap. I see?
 
Great fun to get replies!

Tubelab, I'll check out the Trancendar transformers. I don't have much cash, though. Are the budget ones the same as he used to sell?

Eli D., I didn't think through the gain structure. My current amp has an AV of 18 but the pot (log) has never gone beyond one half (with a low level input). I'll have to calculate whether this amp would have enough gain. I'll empircally calculate the gain I am actually using in the current setup and double that for a target. As far as the driver, I've got a bunch of small signal tubes that I haven't gone through. I will look for some 12av7's, which I probably have. I also have a bunch of of 'sl7 types.

Does anyone have thoughts on pros/cons of fixed battery biasing the output stage? This is fairly common in OLD schematics, but I never read of anyone doing this today.
 
Does anyone have thoughts on pros/cons of fixed battery biasing the output stage? This is fairly common in OLD schematics, but I never read of anyone doing this today.



Lots of tube amps use a C- bias supply attached to the O/P tubes. A SS rectifier instead of a battery is employed. A battery should be fine. The "gotcha" with grid bias and DHTs is the possibility of runaway. You can deal with that issue by using combination bias. A RC bias network insufficient to produce the entire negative voltage goes between the the filament and ground. A "C" battery makes up the difference. Use a WELL ventilated 1% metal film part whose wattage rating is CLOSE to the dissipation you expect. If the tube runs away for any reason, the cathode resistor will "fry" and current flow will cease. The carefully sized cathode resistor also serves as a fusible link.

Have you given thought to the DC heating issue? Even with superior nulling via a pot., hum levels are likely to be unacceptably high.
 
Administrator
Joined 2004
Paid Member
Fixed link:

www.machmat.com/sales/kits/vccs.htm

This will work just fine.. Machmat has a lot of cool stuff.. Another option would be to get a set of Ronan regulator boards and build them yourself.

You can also build a simple current source using LT1085 - just about everything you need to know is in the Linear Tech app note.

Yes you will need a lot of filter capacitance, but 16V electrolytics are pretty small, and very cheap. I use MBR340 schottky diodes in either a bridge configuration or standard fullwave center tap with an 8V (or 16VCT) transformer for 5V tubes and a 9V (or 18VCT) transformer for 6.3V tubes. First filter cap is typically 10000uF/16V. I usually stick a 2200uF on the output, for additional ripple reduction, although some purists argue that this defeats the whole purpose of constant current sources. I'm mainly interested in the constant current operation. Simplicity itself.

The other less simple implementation is to use a a voltage regulator driving a constant current source in which case you definitely can omit that last cap I mentioned in the previous paragraph. Use a 10V transformer, set the voltage regulator for about 9V out, and the CCS for your required filament current. (Using LT1085)

In either case one is required for each channel. Design them to float and do the cathode grounding in the amplifier chassis assuming two chassis design.
 
To prevent runaway, would there be any drawbacks to just adding a real fuse between the cathode and ground?


You are going to have difficulty getting a fuse of the exact rating required. Also, to avoid purely mechanical connections, fuses with pigtails are needed.

Combination bias has a lot going for it, including: sounding good and stabilizing the DHT's operating point.
 
I work at a university and have access to an electronics shop with many values of fuses (all govt surplus so I wouldn't be stealing), so I don't think getting the precise value would be a problem. However, stability of the OP is a consideration and a good reason to have something under the cathode. However, the whole point of the battery is to avoid the bypass capacitor around the cathode resistor. In my small experience they always sound worse than battery fixed bias. I guess it would be possible to have an unbypassed smallish resistor (just enough to keep the tube from running away) and have the battery make up the remainder of the bias. This would raise the effective plate resistance of the tube but maybe not enough to harm the loadline. It would also add some linearizing df.


BTW, I checked on the 12av7s and couldn't find many miniature triodes and thinking through it I the 6sn7-2a3 may not have enough gain for one of my sources. I have lots of octal tubes (including many 6sl7s) and several four pin types, including some 26s. I haven't been crazy about 6sl7s, but thinking back, I have probably never loaded them sufficiently. Maybe one of these with a CCS plate load? Or how about a 26-26-6b4g amp?
 
A common cathode 6L6 DC coupled to a 6SN7 cathode follower will give you plenty of gain and good drive for the DHT "final". Please observe that only 1 coupling cap. (between the CF and "final") is used and that the amp, as a whole, remains non-inverting. Another advantage to the topology is that the DHT's grid leak resistor can be relatively small, which is extra insurance against runaway from contact biasing effects.

Your remarks about an unbypassed cathode resistor are spot on. Even a "medium" value resistor is OK, if you use O/P "iron" with a 3 KOhm primary.
 
In looking for a driver for the 6b4g, I came across the 6c45pi. I guess this tube is commonly used as a 2a3 driver. Incorporating it into the previous circuit with some of the suggested changes, I came up with another circuit. Then I realized that the 6c45pi by itself loaded with the Edcor 5K:16 trannies connected to 8 ohm speakers would give 1.6 watts before clipping when driven by a 2v CD player. To increase the output, I could deepen the bias and shift the plate voltage, but I was trying to keep close to the design centers that I read published on this tube (40mA@150). I think that's all I need. I have played a 200 Hz sine wave into my current amp and turned it up until subjectively loud and I am at .6v into 8ohm speakers. My speakers must be sensitive and I don't listen to music loud. The subs take care of the bass.

So, the project started as a way to hear a DHT amp simply because I have never heard one before. Now I am considering building a Spud because I have never heard one before. Has anyone compared the two configurations? I am intrigued by such a simple circuit. How badly does the 6c45pi like to oscillate?
 

Attachments

  • battery bias spud.jpg
    battery bias spud.jpg
    20 KB · Views: 146
Sorry for the thread drift with the spud idea. Back to the original project to build an amp with minimal cost. I took Eli D.'s advice and looked more carefully through my stash. Turns out that I do have some 12av7s. That's a nice tube on paper. Here is the ammended schematic incorporating everyone's suggestions. I have put an IC CCS load on the 12AV7. I appreciate the feedback.
 

Attachments

  • battery bias 6b4g.jpg
    battery bias 6b4g.jpg
    27.7 KB · Views: 133
Administrator
Joined 2004
Paid Member
That will definitely work, however I was thinking that you ought to put it in the plate circuit, that way if it opens you won't have hazardous voltages in the cathode circuit.

I'm also probably needlessly phobic about the audio current flowing through the fuse in the cathode circuit and the small, but sometimes audible nonlinearity that can result.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.