leedsradio.com Blog

There is a currently a problem with the links. I hope to have it fixed soon.

Looks like it’s up and running now. If you experience any issues please contact me.

Isn’t it rich?

Virtually every week I hear someone say, “I am building a clone of a Fairchild 670, or Pultec EQP1, or [insert name of famous piece of gear here]. Clone is just a euphemism for copy. That doesn’t sound so creative, does it? Copying the Fairchild 670. Friends won’t be quite so impressed. No loud talk of making a copy while drinking sub-standard bourbon, in a bar in the trendy part of town.

About 40 years ago, I sold my National NC-33 receiver to my friend G. He used the radio for a while and then sold it to a person, who should remain nameless. When I asked G. how his customer was enjoying the radio, he told me that he wasn’t. The buyer decided he wanted to upgrade the internal appearance of this perfectly functioning piece of equipment. To this end, he removed every resistor, thinking that all resistors were of exactly the same function and rating. This in spite of the obvious physical differences, including color code and size. He also did not bother to record where each resistor was removed from.

Something this stupid doesn’t happen often but people do often stop by Leeds with a big list of parts they want to buy to replace virtually every passive component, hoping that this will fix their non-functioning piece of gear. I have two important things to tell you about that. First: Replacing all the components is not troubleshooting. It’s stupid. People who want to do this have absolutely no idea how anything works. Secondly, replacing all the components increases the possibility of collateral damage. Vintage gear can be fragile and component removal in hard-wired chassis is difficult. The leads may have two or more wraps around the terminals they are on. Removing the solder and then unwrapping the leads while they still may be partially attached takes skill and patience.

Want to fix something electronic?

Baby step: Learn to read schematics. Learn every single symbol for every single component of the gear you are working on.

If you can’t do this go back to texting your friends or watch some cartoons.

Next step: Understand the circuit. It’s best if you start off knowing what every single component does. If that’s not possible you have to at least know exactly how each block works and what it does. If you can’t do this, you have very little chance of fixing the device.

Next: Unless you are really familar with the piece you need documentation. Schematic, manual, voltage chart, etc. Don’t have those and you’re not familiar with the piece of equipment? Then you have very little chance of fixing the device.

Next: Fix the most obvious problems: burnt resistors, missing parts, damaged parts, and any safety-related issue. Also check that it has the correct fuse. You can then power it up and make some measurements. Unfortunately this means you still need knowledge and experience, especially if you don’t have documentation.

Next: You powered the thing up and it still doesn’t work? You better have a scope. Don’t have one? Unless the thing is simple, it’s not going to be easy. If it is simple a DVM can work. And whatever piece of test equipment you use, you better know how to use it.

To learn all this takes time and patience. It’s the arrogance of the internet age that makes people think that they can do something without learning it first. I was once at a breakfast buffet where the chef at the omelette station could make two perfect omelettes, in two different pans at the same time. It was a ballet. I thought to myself that if I made two omelettes at a time every day for the next ten years that I couldn’t do that as well as he did. Nothing that matters is easy to learn. When you see people that make it look easy it’s because they are enjoying learning. Their gray matter, their motor skills, or strength is still working hard.

I love reading the gushing descriptions about how tubes are supposed to sound.

“Blow Hard brand 12AX7’s have chocolatey mid-range, solid bass and extended highs…”

Really? I have more tubes than most people will have in 100 lifetimes and I have never heard a tube make any sound at all. Yes, I know what they are trying to say but they are misleading people with these descriptions of

I recently received an email from someone who built adapters to use 7F7 loctal tubes in place of 6SL7GT’s. He said that he didn’t like the sound of the 7F7 because the gain was too low. This was news to me because RCA’s RC-22 tube manual states:

For maximum ratings, typical operation as A1 amplifier, and curves refer to refer to glass-octal type 6SL7GT

A quick check in Tube Test Conditions for Hickok Cardmatic Tube Testers, published by Hickok in 1962, shows that they are tested with exactly the same electrode voltages and have the same full-scale gm. The Triplett 3444A, however, does test these with different parameters. The 7F7 is tested in a region of the curves where the gm is lower than the conditions where 6SL7GT is tested. This is an anomaly and gives beginners the impression that the 7F7 has lower gm. Gm is not a fixed number for any and all conditions. If it were the characteristic curves would in fact be straight lines.

In RCA tube manuals you can find nomographs to estimate gm, rp, Ib, etc at voltages/currents other than those listed as the test conditions.

A Sylvania JAN 7F7 loctal with its close relative, an RCA JAN 6SL7GT

Anyone who has ever been out to Leeds knows I love bicycles. An acquaintance has been searching for a new bike and has been asking me for advice. On several occasions he has found something within his price range. But each time he told me he intended to low-ball the seller by as much as 50 percent. The seller

Some years ago I received a phone call: “Hello. My friend says you’re an expert on setting up moving coil phono cartridges. I have a problem: when playing piano records, during crescendos there is distortion on one channel only.” This was any easy one. Many cartridge manufacturers are optimistic in their minimum tracking force ratings. I told this person that he should start from the beginning and set the tracking force 0.1 to 0.2 grams higher than the minimum, adjusting the anti-skate accordingly. He replied, “No. I think it’s cartridge loading.” Very unlikely. Cartridge loading affects transient and frequency response and to a lesser degree output. I explained that I was quite certain that what I was suggesting would fix the problem and that if it didn’t the cost of trying would be zero. Didn’t your friend say I was an expert?
He hung up.

Last week a man appeared at the door. He said he was having a problem with excessive plate current on his Dyna ST-70, a piece of equipment I am so familiar with that I can draw almost the entire schematic from memory. He kept insisting that the “bias” was too high. I told him he had insufficient bias. He replied “But the plate current is too high!” He has confused grid bias (often called bias for short) and the plate/cathode current that bias controls. He’s measuring the voltage across the 15.6 ohm cathode resistor that is brought out to the front-mounted octal test socket.

I received a follow-up email from him this morning saying “…it turns out bias level is not the problem…simply tested the bias range available without the power tubes installed…plenty of negative bias available.”

But he hasn’t actually tried it. He hasn’t tried it.

And if the bias voltage wasn’t a problem, the bias this morning is exactly what it was a week ago. If he puts the tubes in they will still have excessive plate current. This person simply does not understand that if everything is right with his ST-70 then the tubes he has are not good, and if the tubes are good there is something wrong with his ST-70. High plate current is a serious issue on any amp but particularly on amplifiers with smallish power supplies like ST-70’s. The transformers already run hot and the single GZ-34 rectifier is well stressed; the four EL-34’s alone take 200 ma of the GZ-34’s rated 250 ma.

And he absolutely refuses to supply any numbers as to what he feels is “plenty.” I’ve used more voltmeters more times than I’d like to remember and I’ve worked on thousands of pieces of gear. I’ve never seen a voltmeter with “plenty” on its scale. They have numbers. It’s numbers that make technical conversations possible and meaningful. (right : a typical analog meter that can be used to measure bias. Where’s the “plenty” indicator?)

Meaningful. That’s crux of the matter. People who call or stop by and play these games are not looking for meaning or a solution to a problem. They want to engage you. Ensnare you in their world. And it’s a topsy-turvy world. A world where everything is relative, there are no facts and no one knows more than anyone else. It’s also a world where they think I am supposed to have unlimited time to answer their psychological/spiritual/emotional problems disguised as technical queries. I am not supposed to get paid because according to them, my job is too much fun, and fun jobs don’t pay. Please don’t misunderstand. I love giving free technical advice as time permits but my requirement is that conversation is fruitful and enjoyable.

I will never forget the first time I saw metal tubes; it was a pair of 6L6’s in a Bogen amplifier from the early 1950’s. Metal tubes were introduced by RCA in a Technical Bulletin in 1935. The first metals were: 5Z4 rectifier, 6A8 pentagrid converter, 6C5 triode, 6F5 hi-mu triode, 6F6 power pentode, 6H6 dual diode, 6J7 pentode, 6K7 remote cut-off pentode (called a super control pentode in the bulletin). Metal tubes were also the first vacuum tubes to have an octal base.

Most people currently building tube gear don’t like using metal tubes. This is strange because they seem to always like older versus newer. It’s almost as if they think metal tubes as lesser just because you can’t see the filament glow. That’s too bad because the earliest metal tubes are superior in construction and quality compared to the older ST enveloped 4, 5, 6 and 7 pin tubes produced earlier. Sorry folks, the mystique of an ST envelope and nostalgia don

Someone ordered an Amphenol 86CP12 plug and backshell a week ago. Amphenol made a whole series of these to mate with tube & connector sockets. They were available as 4 pin, 5 pin, 6 pin, 7 pin, octal, 9 pin, 11 pin, and 12 . The great thing about them were that they all mounted in the same size hole, with the same ring, or on the same backshell. They are simple connectors designed to do their job without you having to think about them. They are also ridiculously easy to solder and assemble. The 9 pin was used as the AC line connector on the $3,000, Collins 51-S1 communications receiver and also on some Hammond Organs. The 11 pin is used as the power connector on the R390A, triple conversion receiver.

More than 50,000 390A’s were made for the US government between 1958 and 1984. The first ones cost $1500 and the last $38,000. The 86CP8, octal plug is used on the McIntosh C8 phono preamplifier’s umbilical cord. The 86 series males and their 78 female mates are used on more pieces of equipment than you can count and people still use them today when home building when a simple, relatively rugged connector is needed. Not a week goes by when I don’t sell a least one from the series.

Back to my customer. He ordered the connector by part number and received it yesterday. His email to me states:

“It is not anywhere near the quality that we need. I’d like to return it. “

I reminded him by email that he had ordered it by part number and he replied:

“…cosmetics and mechanics of the housing are very low quality and not compatible with our high end equipment. It has a very poor cosmetic appearance and we just can’t use it on our products…”

Now I really had to find out what unbelievable “high end” equipment he was making!

Answer: He’s rebuilding consumer grade Technics reel-to-reel tape decks from the 1970-80’s. So, the 86CP12 is not good enough for a late 1970’s, 240,000 yen (Approx $2,200 back then) but the series is good enough for some of the finest tube communication gear of all time? And what’s he charging for these pieces of “high end”? Nearly $5,000.

It’s late in the game for tape machines. You can buy a Studer 810A for less than half that. One of my customers got a Stevens, with interchangeable head stacks, that originally cost nearly $80,000 new for about $5,000. Technics machines are home machines. I’ve never seen one in a real recording studio. I might have seen something like that in an episode of the Partridge Family. Technics decks have unbalanced inputs using RCA jacks. Nothing is less high end than an RCA jack. You can gold and rhodium plate them all you like they are as connectors go garbage. The cheapest Switchcraft XLR is infinitely better than the most expensive RCA jack

It’s the end times. It’s no longer enough that a part is reliable, fairly priced and above all works. Function is out the window. I guess I could take the backshells, removed the finish, burnish the metal, polish it and then gold plate it, and charge 10 times the price. Then it would be “high end.”

Clockwise from the left: 4 pin, 5 pin, 6 pin, 7 pin, 8 pin, 9 pin, 11 pin, 12 pin. In the center two style of backshells

This tube is more than 9″ (23 cm) tall. Maximum plate dissipation is 1000 watts. Max plate voltage rating is 30,000 volts. The 4PR1000B is pulse rated version of its cousin the 4-1000A.
Filament takes about 21 amperes at 7.5 volts. Filament power alone is so high that you can damage the glass seals if you operate it without forced air cooling