BA-5000 Rework

So, I've been trying to digest what I've been reading about matching paralleled output transistors, and can see where opinions favor closely matching Vbe, hfe, or both, with some indicating that matching should be done over a range of currents as well as quiescent current. My guess is the chances seem small that the Sanken devices I have on hand can pass very stringent matching parameters, and I will end up getting a large batch MJ21193/4 outputs to match from to complete this project. I do still hope to understand better how the 908/909 devices are not suitable to commingle. This is the 908 datasheet:

2SA908.jpg

Acknowledging the responses that advise against using both 908 and 909 together, I am curious about how the specs indicated in the datasheet would make this inadvisable. Other than the Vceo and Vcbo having different maximum specs, all other parameters are identical amongst the 907/907/909 family. Does a different maximum voltage rating change the response curves for all the other parameters even though the ultimate specs for those parameters are the same? I'm pretty dense about a lot of this stuff and am not questioning those who are obviously more knowledgeable; I'm just hoping to gain a better understanding.

One interesting thing I notice from this data sheet is that Hfe is only indicated as a minimum, while the service manual notes different gain groupings. Maybe there are other pages of the datasheet that weren't included in the pdf that I found...

Again, thanks for any illumination on this subject-
 
One thing you should be certain is HFE. This has to be within 2 units. Now vbe multiplier is a whole circuit and must be addressed like one. If you have units 908's and 909's that can be placed together , regarding HFE, go ahead , otherwise don't . As simple as this.
 
One thing you should be certain is HFE. This has to be within 2 units.

The more I read about matching, the more questions I have...

It seems that it is a good idea to match both Vbe and hfe, but not everyone agrees on the parameters under which they are matched. And also, whether it is DC or AC gain that it is more important to match and how closely. While I appreciate the suggestion of a 2 unit match on hfe, this brings up the question of the context of measurement, and whether an absolute value such as 2 might not be strictly relevant to a dynamic spec like hfe.

If you consider the specs for the 908, would the 2 unit variance be in relation to the minimum hfe of 30? And the spec is for DC gain, and many contend that AC gain is more relevant to the behavior in the circuit. But either way, I am beginning to look at these measurements in terms of percentages rather than absolute value. In the case of a 2 unit match at 30hfe, this is the equivalent of +/- 7%, or a 14% range. This is somewhat in line with the suggestion here of a 10% match:

http://sound.whsites.net/transistor-matching.htm

With both Vbe and hfe varying with current, it seems that looking at percentage variance over several levels of current, including quiescent and close to maximum, will give the best indication of matching. In this specific instance, the BA5k calls for a quiescent value of 180ma. Being divided among the 4 paralleled transistors, the quiescent current level for a single device is 45ma. At the high end of the range, with 53V rails, at maximum power, the current per device is 1.65 amps. My curve tracer doesn't operate in that kind of range, so I have to do some more sleuthing on how to measure (probably AC) gain at that level, but my preliminary measurements of Vbe show that all of my 908 devices are within that 10% range.

So, more reading and more measurement to be done, but with any luck I will have suitable matching among my 908 and 1585 devices and won't have to look at the 909 and 1586 units...
 
After much forum reading, testing and head scratching, I have finally come up with a plan for the output devices for the BA5k. In addition to my usual molasses-like plodding along, I also suffered a pretty severe case of "paralysis by analysis" as I pondered how to reconcile the optimum choices with being pragmatic.

At some point, the eureka moment occurred when I realized that, based upon the instructions in the service manual, the Sansui engineers must have heavily relied upon the emitter resistors in this design to force current sharing among the output devices. The specification for replacing any given output device was simply that the new device be in the same gain grouping as the original, and it had become obvious that the gain parameters within these groupings was certainly not stringent and that I should allow myself some fairly wide latitude in matching.

Initially I had leaned toward using contemporary MJ21193/4 devices, but this project is ultimately only for my use and I felt I would take a chance with reliability and try to use the Sanken devices that I had on hand. Using the rig shown in an earlier post, modified somewhat, I eventually devised a set of matching tests that I felt I could apply consistently. This is the final version, with two different pots to allow switching between lower and higher resistance between base and ground resulting in 45ma and 2A levels at 50v.

IMG_1354.JPG

IMG_1358.JPG


There seems to be some consensus across many diy forums that Vbe at quiescent current is the best method of matching parallel outputs, and with the fact that I did not feel I could reliably measure Vbe at higher currents, this has been the primary value that I am comparing. In this case, I tested the transistors at 50V and 45ma of collector current. However, the first test I would run was for hfe at 50V and 2A collector current, allowing the device to heat up somewhat. The collector current would stabilize within 15-20 seconds and I would take the reading of base current, and then switch between the two different potentiometers on the rig to drop the collector current to the 45ma level. The Vbe is very temperature sensitive, so once I had the collector current dialed in to approximately 45ma, I would power down for 3-4 seconds to let it cool and then bring it back on line, do a final tweak for 45ma, then wait 30 seconds to read the Vbe. At that point there is still some gradual drift at the 4th decimal point but the millivolt level is fairly stable.


With this approach, I measured all of the 2sa908/9 and 2sc1585/6 devices that I have. Based upon the best matching I could find, my plan is to use the 908/1585 combination on one channel and the 909/1586 combination on the other. Since there is as much variance in Vbe within specific gain groupings of the 908's and 909's as there is between the 908 and 909, I am not too concerned about channel mismatch. Same for the 2sc devices. With this approach, I can have fairly stringent matching within a channel, with the Vbe specs within 2-3% and the hfe within 10%. And considering the wildly variable Vbe and hfe levels I found within any given gain grouping of these devices, if the service manual prescribes only maintaining the same gain grouping, I am probably somewhat ahead of the game. Not perfect, but it seems that things will fall within a range that would have been acceptable at the time it was built. And if it turns out that I have been wildly optimistic, I can start over with contemporary MJ devices.



Now, it will be time to check the driver transistors...
 
Some progress made...

For the driver transistors, I have used transistors that are badged as GAS118C and GAS119C (Great American Sound), but they are, according to internal documents from GAS, the Sanken 747A and 1116A. I had on hand a dozen or so of each, so I was able to gain match them much more closely than if I had used the drivers that came out of the original and parts amps.

This is the heatsink now populated with all TO-3 devices:

IMG_1377.JPG

Because I could envision that using mica and thermal grease on all 20 of these devices would end up with much swearing and thermal grease in many places where I did not want it, I opted to go with silpad insulators. While it is true that a mica/grease interface would be the optimum choice, these particular pads at least are the Bergquist K-10 pads that I had read were deemed acceptable by Echowars.

Some extra long standoffs temporarily screwed in to the heatsink allowed me to work with it safely elevated to avoid any damage to the connector boards:

IMG_1379.JPG

And finally, the heatsink safely secured to the chassis:

IMG_1380.JPG

This thing is starting to get heavy!
 
Did you check each transistor case to ground to be absolutely sure that they all are isolated from the heat sink?, gads that's a lot of outputs to install.
-Lee
 
Did you check each transistor case to ground to be absolutely sure that they all are isolated from the heat sink?

I had not, but that is a good idea, and I will certainly check them all before proceeding. With the silpads, I was particularly careful that the heatsink surface was smooth and clean, and that they were all torqued evenly to 6 in/lbs.
 
First order of business today was to check the isolation of the the 20 TO-3 devices that were recently installed. All checked fine, and a thanks again to Overundr1 for pointing out the wisdom of that step.

I had planned to do more wiring in the rear sections, with installation of the fan's thermostat and the PTH487A thermistor onto the heatsink, but I had neglected to order a new 120ohm resistor for the fan circuit back when I was collecting parts, so I moved onto the front end.

This is the chassis with most of the wiring threaded in and ready to receive the front panel:

IMG_1382.JPG

And the front panel with the meters and the buffer board installed:

IMG_1383.JPG

Fingers crossed that those big meters will be operational...

It's a little discouraging to see, as I go through reassembly, just how much rust and corrosion afflicts the panels and chassis parts. Both the amp being restored and the parts donor must have really been abused. At least now the internals may not be pretty, but they will at least be clean.
 
+1 - it's looking very good indeed. ;)

I found re-mounting the front panel, while not allowing the meter plugs to come off was a bit of a nuisance, it seemed too difficult to plug them on after the front panel was mounted - I forget exactly why. You'll have fun with that if my experiences are anything to go by :)
 
Are you planning to install the output fuse modification that Trnsformer was discussing a while back? I would be very unhappy if I had to remove and rework all those TO-3's.
 
Did a little noodling with the cooling fan, and at this point, my plan is to replace the existing fan with a 240V unit to reduce the speed and noise. I put the existing fan, Rotron Centaur 115V model, on the variac and even with the voltage dialed down into the 90's this thing is quite loud. I happen to have a 240V rotron that snaps right into place, and when run at 120V, drawing 45ma, still pushes a noticeable stream of air and is extremely quiet. I am using a 60 degree thermostat in place of the stock 70 degree unit, but even with that I am hopeful that the fan coming on would be a rare occurrence.

The soft start relay went back in today (coil tests fine and contacts seem pretty clean), and the front panel went back on:

IMG_1387.JPG

It's starting to look a little more like an amp...

Other than the line level outputs to the driver boards the buffer board is wired up:

IMG_1386.JPG

Also ran the feeds from the speaker terminals to the meter drivers:

IMG_1388.JPG

I'm a little curious why the drive voltage for the meters is pulled off of the 4 ohm taps of the output transformers since 4ohm speakers were not as common then as they are now...
 
From the speaker terminals, so the output taps on the transformers. But what I just noticed, making it even more curious, is that the schematic indicates the feeds to the meter driver boards coming off the negative speaker terminal and the 4 ohm (output transformer orange wire), which is how this was originally wired, but the meter adjustment indicates the 0db adjustment on the meter while measuring 49V at the 8 ohm tap.
 
Back
Top Bottom