Transformers for Tube Amps

Try Antique Radio Supply: http://www.tubesandmore.com/

I just threw some out. Maybe go to a junk yard and buy an old radio or TV set.

Try looking for surplus dealers like Fair Radio Sales in Ohio, it may take some digging on the web but there are sources of surplus transformers at good prices, new ones are rather expensive. You may not find a single transformer with all the voltages/currents you need so you very well may have to use more than one transformer to meet your needs. Antique Electronics is a good source but limited in the number of transformers available. They can special order Hammond transformers, so can Mouser Electronics, Allied and Newark, check around as there is some difference in pricing. I'd try the surplus route first.

try: Hammond Transformers, Waterloo, Ontario, Canada

they used to manufacture many audio transformers, and can custom wind just about anything.

Hi,

It would be helpful to know what output power you are thinking of. If your power amp is say 50Watts RMS then the valve(tube) line up is probably going to be with KT88 output valves ECC 82 driver ECC83 phase splitter and EF86 pre-am stage. This puts you into the realms of the MARSHALL PA type amp and parts for those are available world wide.

Hope this helps.

Thanks for the comments. I have found transformers on eBay including Fender transformers for the famous amp Fender Deluxe 57 5E3 which was one of the first Guitar amps built. I am building a clone right now and have the PT & OT and all the tubes.

Regards,

maveric_manic

Hi MM,

Good to hear of your success with the transfprmers. Just one thing about the Fender range (and they were not alone) is the hum level. They could be really noisy and the hum could be rather noticeable. I experimented with isolating the pre-amp end physically by separating the chassis with some improvement. however my best result was achieved by running the heaters of the preamp section on DC! To to this I took the 6.3 vAC into a bridge rectifier then used a simple RC network to smooth and reduce the output voltage I also created a centre tap effect using a 100 Ohm pot with the slider grounded; by moving this around the mid position the best (quietest) performance was achieved. I got S/hum signal levels better than 150:1 with this set up.

Let me know how you get on.

Good luck.

Thanks for this very good info. I have been doing research on vintage amp problems like hum. I am using a 100 ohm divider for the 6.3 VAC to chassis ground to simulate a 6.3 VCT. And also keep the transformers away from the preamp section. Use twisted pair wire in phase with "all" tubes including Push-Pulls. A single point chasis ground large guage wire. Shielded input coax wire. I have not seen any body use DC for heaters except +12V in series with a 12AX7 audiophile preamp using DC/DC converter for filament and plate (no transformers) except for the hand made 2 winding coil for the DC/DC converter which the ferrite part is obsolete and I could not find a replacement. So I decided to go the traditional route. Using a full wave rectifier and RC does this drop the 6.3 VDC voltage due to voltage drop across the diodes and also the resistor? Or maybe the fillament voltage is not so critical to be exactly 6.3 V like the DC/DC +12 VDC instead of 12.6 VDC. I have not read anything about fillament voltage tollerence but plate voltage I have read can vary as long as you do not exceed Max. ratings including voltage spikes on the plate supply. Thanks I may do this and I checked my tube manual for 6.3 V fillament can be AC or DC but have not seen DC used except for the +12 V example I talked about but I will consider this.

Funny, with transistors you get white noise and tubes hum 60Hz and 120Hz (from the plate filtering or lack of proper filtering in vintage amps). Thanks again.

In critical, low noise circuitry, DC filament power is indeed common but not very common in consumer amplifiers. The voltage tolerance for filaments is +/- 5% for best life. Generally, the higher filament voltage (12.6, if available) is used for DC since it requires half the current. If you are only powering preamp tubes, which are usually of lower current ratings, you may need to put a dropping resistor in series with the rectified/filtered DC. Exceeding the 5% voltage tolerance will rapidly decrease tube life, the higher you go or lower, depending on circumstances.

Long term, large quantity, studies were done on tubes back in the 1950s for all of the important characteristics of tube operation and life span, it is much too long to go into details here. There are a great many misconceptions about tube operation in print and online, mostly due to the lack of technical information from the past and the general lack of knowledge now. Much has been said, based on incorrect observation and assumption, often by honest mistake, and taken at face value. Curiously, it has been very difficult to get the accurate information published.

You will either have to use a bridge rectifier with a 6.3 volt winding or a full-wave, center tapped 12.6 volt winding, I would go with the 12.6 volt winding which will require a current rating of about 1.2 times the filament load, a bridge will require 1.8-2.0 times the current load. Depending on the voltage rating of your transformer and line voltage, you may end up with output voltage (under load) outside the 5% limit during line voltage fluctuations. A resistor may only partially correct this. If your line voltage is relatively stable, you may be home free.

You can construct your DC supply, connect it to your tubes and use a Variac to simulate line variance, the usual line voltage tolerance is about +/-5% but it can be worse than that. If your line voltage varies too much, you will have to construct a voltage regulator circuit, such as a LM317, to power your tubes.

As to your mention of noise, tubes have two primary noise sources, sometimes called shot noise (and other names) and thermal noise. Thermal is the dominant noise above the corner frequency, below which, the shot noise becomes dominant. The shot noise is due to the variation in electrons hitting the plate. If amplified sufficiently, it can sound similar to rain falling or a 'rushing' water sound. If you are hearing shot noise, it is a sure sign that you are trying to run the stage at too high gain.

Hum is mostly the coupling of AC through inter-electrode capacitance in the tube from the filament and in some cases, radiated AC pickup from the transformer and even outside sources, including the cabling and ground loops. Unless you are working with very low level sources such as low impedance microphones or tape heads, careful layout will usually prevent any significant AC from getting into the signal chain. As you mentioned, a center tapped filament winding and a balance pot very often clears up hum problems.

Hi EW,

Good to get your learned thoughts on this subject. I started messing about with small amps in the early 60s. graduating to larger PA kit and building them as a source of spare time income! The problem of rampant distortion levels and horrible hum levels was always an issue. Separating the various bits and pieces on the bench whilst monitoring the residual noise level showed the way forward; but it only needed a change of transformer or chassis material to upset things. Nevertheless I did manage to construct a few half decent performers in Hi-Fi terms and still have one to this day.

I alwys think that one day I might have another go at making a decent pair; but things seem to get in the way.

Good to hear from you all the same.

Hello Massey....

Thank you for the compliment. While good engineering skills are important, I think that a little bit of 'art' enters into it as well. I also cut my teeth on tubes before moving on to solid-state but I still have kept both sides active. There is much to be said for a good tube amplifier. There is also much to be said about all the nonsense that accompanies tubes these days. The most curious thing is the unwillingness of many to accept tube fact over tube hype.

I quite agree with you, experimentation is a very good way to develop one's skills but one must also realize that it can also be filled with 'traps' and must be accompanied with good research. That can be difficult today as there is a huge amount of nonsense circulating on the web and many people have no way of determining which is accurate and which isn't.

I still use tube amps in my stereo (also have solid-state) and I've got another amp on the project board too, it's been there for quite some time and probably will be for some time to come, too many irons in the fire.

It's just my opinion but I don't like the 12AX7 for high fidelity audio, it really isn't a very linear tube. I prefer to stick with octals such as the 6J5, 6SL7 and 6SN7 types for pre-amps. The 12AX7 is very popular with guitarists because it can be made to distort very easily and they love the effects, which is great for them but lousy for audio. One reason I think people pick the 12AX7 is because they think they need, or get, very high gains from it, in practice, they don't, about 60 to 65 is the top end of gain and that is with a rather high plate load and restricted bandwidth. Higher gain can be realized of course but bandwidth must be very restricted to get anything approaching linear response. The main reason for the 12AX7's existence was manufacturers were wanting smaller chassis and lower costs, it had little to do with audio specs, it wasn't a better audio tube. Yes, it has found its way into a lot of applications but for varied reasons, fidelity not withstanding. It is not the easiest tube to work with.

Hi mm,

Good to hear from you. I experimented with heater voltage as I feared that over voltage would not only risk a burn out but might put the 'innards' under thermal stress. Funnily enough I found that at higher voltages there was an increase in microphony so perhaps things were getting a bit stressed inside. I found that at 5VDC that warm-up time was not compromised and the output seemed smoother.

As my mains transformer was a real big brute I was not worried about losing a few VAs in the input to the rectifier and over all the improvement in the S:N ratio at about -100dB (as against a very variable one) was worth all the time and bother that went into messing about on the bench for weeks!!

What do you have to pay for a good pair of KT88s ?

Keep in touch.