MAXSON CA-1589 Tube Limiter and Tubes

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
> get the frequency response to not roll off below 300Hz.

You noticed that too.

There is a trick of using a steep bass-cut between the vari-gain stage and the next stage, to reduce thump and clipping-time. This is compensated with a bass-boost somewhere else.

But I don't see any bass-boost, so I agree: this is a communication speech processor. Perhaps useful for many musical chores, lead instruments and voices.

I wonder if it will be stable with less bass-cut. But surely it didn't motorboat in the original use. So something has gone bad.

> Not sure why you think 6BA6 will be better as it is usually only good as a Mu tube when run as a triode

6BA6 was intended for use as a pentode in variable-gain amplifiers. Warm-up your grandmother's kitchen radio, turn from a strong local station to a weak distant station, it nearly equalizes the audio levels (not exact, but 40dB into 10dB). 6BA6 is much better than a 6AU6 when you need to reduce gain 30 or 40dB. This is practical and necessary in radio reception. There are real problems with putting 30+dB gain reduction on good audio.

Both triode and pentode modes are possible. Pentode has an advantage that gain reduction does not involve a significant change in plate resistance: all practical loads are "low" compared to a pentode plate. In a triode, as current and Gm is reduced to reduce gain, Rp goes up which tends to increase gain. Triodes must be loaded in very low impedance (with low output), very high impedance (higher than lowest-current Rp, which is usually impractical), or swung over a VERY wide range of current to get sufficient GR (a popular plan). One advantage of the triode, with resistor loading, is that minor tube mismatch tends to self-correct as Rp interacts with the plate resistor and supply voltage.

In this case: while 6BA6 might be better in some way, I have to believe the Old Men knew their stuff and picked 6AU6 for some very good reason. There was a lot of collected wisdom in the old radio-labs, and plenty of test-gear to keep them honest. It sure was not cost: at the time, 6BA6 was more common and probably cheaper than 6AU6. EVERY 5-tube radio had 6BA6 or equivalent; 6AU6 had wide use but not like 6BA6.
 
The unit as it was with C1,2,3,4 set using .001uf 1kv mylars and using the 6AU6 tubes worked fine, no motor-boating. I thought I'd try the 6BA6 combo as recomened by this other owner of one of these.

The 6BA6's with C1,2,3,4 @ original .001 values had motor-boating or pumping every second or two.

The 6BA6's with C1,2 @ .001uf and C3,4 @ .1uf gave off a faster or rapid motor-boating. This cap combo with the 6AU6's instead worked fine, no motor-boating.

The 6BA6's with C1,2 @ .022uf and C3,4 @ .1uf gave off a very rapid motor-boating or pumping. This same cap combo with 6AU6's gave the same very-rapid motor-boating.


So neither 6AU6 or the 6BA6's liked the C1,2 @ .022uf and C3,4 @ .1uf combo, what have they in common?

I'm going to try waht Joe suggested here shortly tonight.
 
[quote author="tubemonkey35"]waht = what :)[/quote]
Yes, we understand. Those "a" and "h" keys are too close on my keyboard too.

BTW, unlike some other forums, here you can edit your own words even after you post. Taht's sometimes cleaner than a follow-up correction. If you catch it quick, nobody will know.
 
I have not as of yet tried the experiment as my wife is now home.

A friend of mine wonders if the 6BA6 tubes are taking too much voltage causing them to thump.

I guees the experiment if it proves true (removing the 6AL5) then we know it's not that.

Crap, while I'm adding an attack control, why not talk about adding a release too.
 
Ok, I just simply just pulled the 6AL5 out of the socket. I put the 6BA6's back in and I still got rapid motor-boating.

I do not get motor-boating with the 6AU6's while the 6AL5 is pulled.

I guess this rules out the too Fast Attack theory? Where does this point me?

Why is it that both 6BA6 and 6AU6 setup did not like the C1,2 @.022uf and C3,4 @ .1uf combo as they both motor-boated? Being that the 6AU6's worked fine in the (C1,2 @ .001uf and C3,4 @ .1uf) setup while the 6BA6's did not , does this lead me to believe it's not a tube balancing issue?
 
Thanks Jakob, I'll am going to try that. Looks like Late tonight or in the wee morning hours before my wife awakes.
 
I agree that there is a balance problem as well with the 6BA6 but I would get the comp working with the original 6AU6 tubes first or you will be trying to solve to many problems at once. I am sure there are going to need to a few resistor changes to get the 6BA6 stable. I have had this exact problem putting 6BA6 into other Mu Comps.

So far you have proved that the attack is making the 6AU6 motorboat now that the audio at full range.

Does the compressor sound full range now with the 6AU6 and no 6AL5?

Do all the things I mentioned about getting the attack right a previous page to get the comp stable with the 6AU6 with full range audio and then work on getting the 6BA6 to work. Not forget if you are running line level into the comp wire in the 27dB pad.

Joe

www.jlmaudio.com
 
I have it hooked up with the 27db pad in place as well as being set for 600 ohm line input.

I have not done the balance yet, I have a 250k ohm 1/2 watt pot and wonder if that might burn-up in there; I could not find a 1 watt lin pott at the surplus store.

I have also just feng shui'ed my whole studio / house by relocating my control room yesterday. This was my wife's idea and now my console sits in front of a bay window where my dining room used to be. This has slowed me downin getting doing the 6BA6 balance test, however I may get lucky and give it a go tomorrow.
 
I've rebuilt and modded several variations of these communications limiters. I tested all transformers out of circuit (both a Maxson and another manufacturer) and found that only the final repeat coil passed anything like flat audio. Some of the inputs and outputs did okay on the high end, but were severely rolled off on the bottom. I found the same motorboating issues with increasing coupling capacitance, and no amount of additional B+ filtering helped. And I mean insane overkill amounts. I was able to squeeze a little more lows out of some increased coupling, and a little more out of feedback tailoring. But not very much. I think I tried a higher quality input trans (can't recall anymore; maybe making this part up) and it didn't help anything. The output trans has a 60K to 600 ohm stepdown between primary and secondary, and the tertiary winding used for sidechain rectification has a 1:7.3 step up ratio. So it is pretty much impossible to replace with higher quality. You'd be better off starting from scratch than trying to make one of these be full range, or even close.

Face it; these things sound like high quality telephone lines at best. That's all they are meant to do. Neat special effect, if you really want an expensive special effect. I saw one of these limiters top $1700 this spring. You'll be most of the way there by sticking a bandpass filter in front of a regular vari-mu limiter.

I sold Michael Brauer the one he's raved about so much. I believe he modded mine further than I had done.
 
Thank you for the quick reply!
I have a 1589 in for service,and it looks like it has the modifications you mentioned.
Feedback resistor and capacitor from V4A anode to V3B grid 1, and the same for V4B to V3A.
Bigger values on the capacitors,and some alterations of resistor values.
Looks like the gain of the output stage have been reduced.
The varimu tubes is 6BA6. Maybe it is your mod?

It has the mentioned problems with oscillation.
Did you ever try to use "perfectly matched" tubes?

Do you have any notes on the modifications you made?

I am thinking about making a test jig for 2 single 6BA6 tubes,with manually controlled negative control voltage,and use same phase input signal on both tubes. Then I connect a transformer between the plates.
If the tubes have the same characteristic, there will be no sound on the output of the transformer.
If they dont match,there will be sound with amplitude according to the mismatch.
Higher amplitude,bigger mismatch.
Does this seem like a smart way to find mathing tubes,or am I on the wrong track here?
I am going to test the tubes to match the plate current as close as possible before trying to find pairs.

Any comments?

I will of course try the potmeter matching method....


Kåre
 
I'll try to dig out my notes next week when I'm back around. I stuck with the original 6AU6. Sounds like the one you are looking at has further mods. Note that all of mine worked perfectly with any 6AU6 I wanted to try as long as I left them stock, or modified for extra bass by only a little bit.

The problem with matching is that the match will vary depending on the amount of gain reduction at any given moment. So it is an inexact science, but you can get close. I tried a very large pile of tested 6AU6's, using a control in one position and swapping others in and out of the other socket. It will be hard to know what matches in the system unless you rig a dual current measurement system (or galvanometer between the cathodes?) and watch at varying amounts of reduction. Your test method sounds like it will work, but it's still dependent on perfect transformer balance and response at all affected frequencies. May differ from the experience in circuit. The cathode balance pot is still the easiest method, once you know you have good tubes and good resistor balance at the plates. [/i]
 
I have been thinking about new transformers and output tubes in this one..
Replacing the output transformer with Lundahl 1680,and connecting the 600 Ohm output to the existing output transformers 600 ohm output, to get the control voltage..
Of course with the primary windings disconnected.

Then using something like ECL86 for the output.
This way I can use the existing tube sockets,just rewire them..

Any thoughts??

Does anyone have a suggestion to which input transformer to use?

And does anyone have a schematic of the input attenuator?
Would be nice to know what kind of loading of the IP transformer to expect.

Kåre
 
update on an old one. 

I FIND IT REALLY IRRITATING THAT THE FORUM SOFTWARE THINKS I SHOULD BE WARNED ABOUT POSTING IN AN OLD TOPIC, AS IF IT'S HOPELESSLY UNPOPULAR, OR SOMETHING.  WHY THE F&% WOULD I WANT TO START A NEW TOPIC?  ISN'T THAT WHAT'S WRONG WITH MOST FORUMS ANYWAY?  TOO MANY REPETITIVE POSTS, AND PEOPLE COMPLAINING ABOUT INFO SPREAD AROUND ALL OVER THE PLACE?

I stumbled across my notes, and I see that these things measure -11 dB at 100 Hz, and -8 dB at 10K, relative to a flat range at 0 dB from 1K to 3K.  Yes, that's right.  I measured both a Maxson and a Tri-tronics model.

Check it:
http://groupdiy.com/index.php?topic=33552.msg469161#msg469161 
 

Latest posts

Back
Top