Spec-1 ACV-314 pot lessons learned

This is a 4 wafer 22 position SWITCH with staggered resistance steps.

Two wafers have different resistance tapers than the other two wafers.
It is possible that two are log tapers and two are linear tapers - or some close variations working together.

Each wafer is a ceramic substrate with printed resistors. The commutator is a dual fork, with triple sliders to the switch contacts. The tubular spacers and disk shaped commutator/slider holder appear to be nylon.

The shaft is solid, brass and is prevented by the mechanism from inward over travel.

Outward pull is another matter - a dangerous one. The mechanism that retains the shaft in the housing is a staked on square steel plate with a spring / detent roller / rotation travel stop assembly that will not withstand a lot of outward pull before coming apart - the stakes ARE only brass....

Thus one can loose the detents, (sometimes while still preserving the rotation stops) which seems to mess up the first wafer in the stack as well.

The removal of a stuck knob becomes problematic. The knobs are solid aluminum with 1.5mm set screws - and the set screws when over tightened can dig into the brass shaft and displace material - making knob removal very difficult.

Since the knob is solid aluminum, I have successfully used a heat gun to heat expand the knob and reduce the force necessary to remove it.

Next after the detent mechanism is a long stack of wafers, culminating with a round metal disk staked on the end of the shaft. I suspect besides holding the end commutator/slider disk it also acts as a shield.

Undoing the staking is an irreversible step, and my best result to date is to grind down the end of the shaft flush to the disk. That allows removal. The metal disk when appropriately prepared will take solder, as will the brass shaft - thus the last step upon reassembly of the stack is to solder together the formerly staked together disk and shaft.

BUT this has to be done with appropriately large soldering equipment with a large heat reserve, so you can get in and out within a second or so, to avoid harming the nylon. THEN LET IT COOL!!!

Pre "wet" the disk and the shaft (END ONLY)externally with no nylon in contact with it.

The beauty of this approach is that while the staking is a one time deal, the soldering can be repeated carefully.

Also be aware that in the stack are metal shields between the wafers, that have an extremely tight fit to one screw - for conductivity..

Now to the meat of the matter, the resistance taper. Where you see two consecutive identical readings, it is a wide contact and we are on the same position in the string of resistors.

the first column is the panel marking for the position, which is negative dB or attenuation.
the second column is wafers one and two, resistance to "ground" tap in kilo ohms.
the third column is the resistance between the steps of the second column, which add up to 50.4 k ohms
the fourth column is wafers three and four, resistance to "ground" tap in kilo ohms.
the fifth column is the resistance between the steps of the fourth column, which add up to 50.4 k ohms

.......... #1 ................ #2
inf ..... 0.0 ..... 0.0 ..... 0.0 ..... 0.0
-60 .... 0.5 ..... 0.5 ..... 5.2 ..... 5.2
-50 .... 1.1 ..... 0.6 ..... 7.5 ..... 2.3
-42 .... 2.0 ..... 0.9 .... 10.9 .... 3.4
-37 .... 3.2 ..... 1.2 .... 12.3 .... 1.4
-33 .... 4.1 ..... 0.9 .... 15.7 .... 3.4
-30 .... 5.2 ..... 1.1 .... 17.7 .... 2.0
-28 .... 6.6 ..... 1.4 .... 17.7 .... 0.0
-26 .... 6.6 ..... 0.0 .... 22.4 .... 4.7
-24 .... 8.5 ..... 1.9 .... 22.4 .... 0.0
-22 .... 9.6 ..... 1.1 .... 25.2 .... 2.8
-20 .... 12.2 .... 2.6 .... 25.2 .... 0.0
-18 .... 13.8 .... 1.6 .... 28.2 .... 3.0
-16 .... 17.7 .... 3.9 .... 28.2 .... 0.0
-14 .... 17.7 .... 0.0 .... 35.1 .... 6.9
-12 .... 22.5 .... 4.8 .... 35.1 .... 0.0
-10 .... 25.3 .... 2.8 .... 38.7 .... 3.6
-8 ...... 31.6 .... 6.3 .... 38.7 .... 0.0
-6 ...... 35.1 .... 3.5 .... 42.6 .... 3.9
-4 ...... 42.6 .... 7.5 .... 42.6 .... 0.0
-2 ...... 50.4 .... 7.8 .... 42.6 .... 0.0
0 ........ 50.4 .... 0.0 .... 50.4 .... 7.8
.................. ------ ............. ------
................... 50.4 ............... 50.4

Thus to "roll your own", with a four deck 22 (or 23) position switch, these are the resistance values, and switch positions to connect together to make a substitute.

Notice that from the first two wafers to the second two wafers, there is always a resistance change on at least one of the wafers when switching to the next position.


Sorry, no pictures, by the time it occurred to me to share these observations, I wasn't going to disassemble the finally working, and tested assembly. This is NOT for the mechanically disinclined or faint of heart. Considering the current values of Spec-1 units, this is a last resort to recover a busted pot rather than killing a donor unit.


salient details:
330 degree rotation - confirmed!!
15 degree indexing - confirmed!!
22 stop positions.
4 decks/sections
2 somewhat linear (yellow) (red is a linear reference)
2 somewhat log (blue)


here is a graph of the curves:




thus a 4 section 50k pot, with 2 sections linear taper and 2 sections log or audio taper will be a reasonable substitute. If there are no detents on the pot then the pesky problem of the detents not aligning with the markings is eliminated. An ALPS type pot would be perfect.

edit:
There ARE some "pots" coming out of Hong Kong that are 21 position switches with smd resistors and an audio taper, the idea being to have matched resistances between the sections, for tracking. Ganging those wafers with some linear taper wafers looks interesting. One brand has a hex socket indentation that looks useful for further gang coupling.
Other smd stuff is showing up as well, with 23 steps. This is getting interesting!!

edit2: I have just confirmed that the indexing between the steps IS 15 (fifteen) degrees!!! this is a standard value - making the location of a 22 step 330 degree (or 23 steps and 345 degrees) switch MUCH easier: 4p22t or 4p23t.
This beast is starting to twitch... it WILL live!!

Still have some parts of one of these switches off a donor unit Mark. Wafers, cases, wipers, Need It?

Hmmm...... :scratch2:

Your donor unit comes to mind (actually it had BEEN in there all along) and getting a pot together for it. I have a line on 22 position switches and I was going to check on what the degrees per position were before making noise about this.

I've been thinking about MAKING replacements for quite a while, and having a GOOD one to take readings from finally kick started the process.

That's why I calculated the resistance "steps" in the two wafer types...

How many wafers are there? Any broken ones?

This one has a few other busted or bent things as well to work on...

I'm surprised there's as much resistance difference between the channels, considering these pots are calibrated to +/- 0.2dB "step error" and 0.2dB "gang error" :scratch2:

You don't get it.

I had TWO pots, I could have done a duplicate of your picture.

There are a total of eight wafers between the two, and four wafers had one set of readings and the other four wafers had the others set of readings. Thus they DO have two different resistance curves... The different curves were in FIXED positions in the assembly - curve #1 were the first two wafers starting from the volume control knob and going away from the knob, while curve #2 were the last two wafers in the stack, furthest away from the knob.

And I mean that with the 3 1/2 digit ohmmeter I used, they were EXACTLY those numbers. I didn't ALLOW it to switch ranges and confuse the issue with a different calibration for each range.

ALL eight hit 50.4 on the meter, and the SAME numbers same digits on each of the FOUR wafers at EACH and EVERY step. EXTRAORDINARY!!!

To go any further (with my 5 digit lab meter) I want to calibrate the ohmmeter to a very high degree of accuracy first. Then take numbers off of the just one pot that will be available.

I have never (before the spec-1's) run into that level of precision with two randomly selected controls.

So yes, the +/- 0.2db step and +/- 0.2 gang errors are quite plausible and I believe met...

The two curves, and the "steps" work together to achieve the final result at the output of the circuits.

Ah OK I thought the 4 columns were the 4 sections of one control. I get it now :yes:

This is whats left of the pot.

Wow, the two good wafers are the closest ones to the front knob with the first taper. The busted one is the second taper. They would have been useful had I not been able to recover the first wafer in the stack.

Do you think that it's all safer here? Until we find the other two wafers and miscellaneous parts, then that pot can be put back into a unit or another repaired.

Thanks Mark for the post. I now realize how lucky I was when tearing into my own spec one that I did not have the 'stuck on' knob issue that others have run into which very well could destroy the pot if over enthusiastic force was used to remove it. I also feel that not abusing the knob at its stops during cleaning might save those detent stops. Another lesson learned.
-Lee

You bet it would be safer there. How about the rest of the Spec 1 to go with it, I have 3, and I just recapped all 3 and they are doing fine. I don't need a parts unit, and I'm sure you can.

How can I turn that down? OK, you twisted my arm

Have my address easily accessible? or should I send a PM?

Yep, got your address from the last shipment. I WILL get the 850 headed your way also when i get a little caught up. It has been a helluva summer in the oilfield. I believe obama hates the oil industry and is doing everything in his power to nmake our life miserable.

Got the Spec 1 boxed and ready to go. I'll PM the tracking number after I get it.

FedEx, went out yesterday afternoon.
Thanks, Lee

Thanks Lee, I have the tracking number, so you can disappear it from your post.

Just in case somebody with ill intentions reads this. Don't want them to know where a spec is gonna possibly be outside for the taking - since Fedex tends to NOT ring my doorbell, just sneak up to the door, toss and run.

Gotcha

Hey Mark, care to share where to get a pioneer spec 1 replacement volume pot since they are showing up from hong kong?

xXColdShotXx said:

Hey Mark, care to share where to get a pioneer spec 1 replacement volume pot since they are showing up from hong kong?

Click to expand...

I have no idea whatsoever. For real. blank.
Check the stuff above, there are 2 resistance curves. IF they don't match, the volume curve won't match the markings. Might even be annoyingly noticeable.

note:
from top post,

The two curves, and the "steps" work together to achieve the final result at the output of the circuits.

Click to expand...

I have the spec 1 volume out and would like to have it repaired. Would you be interested in cleaning and fixing it? I've asked mattsd multiple times on pm,but he's gotten to where he won't respond to anything lately.

I took the spec 1 volume pot apart and found the front wafer had a bent wiper, straightened it out, cleaned all the metal contacts best I could, put it back together and it's working good now and no static when turning the volume at low level.