Ampex ATR-100 #2

[evm1024]

I've managed to find enough parts to start working on the #2 Ampex ATR-100 parts machine. No photos yet....

I've spent the day working on this deck and thought to give a running monolog of the debug process.

The deck came missing a number of boards which I was able to get from eBay. The big ticket missing items was the tach roller and the capstan roller. I found 2 capstan rollers on eBay which I bought for less than $50 each. The first had the aluminum worn off the glass encoder disk (it forms part of a tach) and the second has the rubber turn to goo. Not all is lost. I have 2 cores now for rebuilding and the Goo'ed one can be used for testing. After cleaning and installing the 2 rollers I powered it up without any cards in it to check the power supply voltages. So far so good. No smoke or odd smells and the voltages are in the ballpark.

Next step was to install the reel servo, capstan serve and transport control boards. And then power it up again. Ah, there is an odor... Power off and start looking board by board. Smell, look, touch. A dual opamp on the reel servo board is hot and the 100 ohm series feed resistor for +12 is burnt.

Pull the opamp and put in a socket, replace the resistor and check some more local parts. The cause was a shorted tant cap. Replace both the tant on the +12 and on the -12.

The + and -12 on the reel servo board come from a voltage regulator circuit on the capstan servo board. The 13 volt zener diode is bad along with the regulation series pass transistor. No 13 volt zener diodes here so install a pot and replace the pass transistor with another one of lower power (all I had but should be more than enough). Test it out and set the voltage for 12 volts.

OK, power up again and no smoke or odors. Progress. Put some scratch tape on it and take the slack up. press the stop button. Odd, no click. Open up the control and discover that the stop microswitch is gone. Parts machine....

Ok use some alligator clips in place of the stop switch. Darn, the reel servo should activate and tension the tape. But nothing. Out comes the logic probe. Press stop (well touch the clips together) and follow the signal. All ok. Ah, the tension arm microswitch is not closed. Looks like a manual adjustment. OK, push the tension arm over manually and try to activate the reel servos again.

Excellent! Reel servos working. Tape is tensioned. Turn capstan manually and the reels move. very good, very good. Tape can shuttle back and forth.

OK, power up the scope. TIme to adjust the capstan tack. There is a glass disk with 1200 lines on it on the capstan and a grating with 2 photocells and LED on a set of small boards that shine through the disk. THe small boards can be moved in and out and rotated to get 2 quadrature signals from the disk. THis tells you the rotation direction and speed of the capstan. Scope them and move the boards for the strongest signal and correct phase relationship between them.

Looks good. Press play. The capstan turns!!! -------> Backwards.

Darn, so close. Check the motor connection. Keyed, no problem there. Time for dinner, eat, grab the manual and computer. Write this and then dig into the manual to try to find out what would make the motor turn backwards.

Good progress for the day. THis deck has 4 track 1/2" heads. I'm thinking of converting it to 1/4" 2 track but keeping the 4 track heads as well.

Regards, Ethan

 

[evm1024]

Solved! Plus Photo

I got some more time and traced the problem down to failed TTL IC. The direction command sourced at a TTL gate and fed 3 other TTL inputs on 2 boards. The signal was either at 0v or 0.12 v. TTL is normally above 2.5 v or so for the high state. Something was pulling it down or it was not being driven properly. Pulling the source output pin on the IC (I replaced that IC first and put in a socket) I was able to see proper TTL logic levels when not driving the loads.

So I checked the logic of each receiver and it appeared that they all were working properly at least with those inputs. Drats!. I just started replacing them one by one and as luck would have it it was the third and last chip (a 7404).

I've got a hoard of TTL and ECL logic chips off in a few boxes so I was able to locate one.

The transport now works! The bearings make noise and the supply reel motor is quite loud on REW or FF but it does work. Plays 15 and 30 ips, rew, ff, stop and even put it into record.

I need to get some replacement microswitches for the control buttons and then some bearings and do a bunch of transport adjustments. From there I need to get the capstan rebuilt. Then I can do a calibration with the 1/2" 4 track head at least till I build a 1/4" 2 track headstack.

Here are photos of the deck with the top plates off and the 4 track headstack next to a Tascam 32 headstack.

--Ethan

 

[cjacek]

Ethan, is that the deck which doesn't use a pinch roller?

I bet my ATR-60 and your ATR-100 could become fast friends. Hey, at least they've got the first three letters in common!

Amazing!

 

[evm1024]

yup

Ethan, is that the deck which doesn't use a pinch roller?

I bet my ATR-60 and your ATR-100 could become fast friends. Hey, at least they've got the first three letters in common!

Amazing!

That is correct. The tape just goes around the capstan (on the right of course) and the tapes tension holds it against the capstan. The tach roller is on the left and counts the time.

--Ethan

 

[evm1024]

Time Flys

Been a while since I've updated this thread..... So much going on elsewhere.

ATR #1 is now 1/2" 2 track with a new (from ATR Services) capstan. This frees up the old capstan which is now in ATR #2. After aligning the capstan encoder everything is good... except that sometimes it gets skitzo.

Looking deeper I see that the takeup arm encoder LED is out. Now you might think that this is just a LED and you wold be right. Except this is a 1975 LED and they just don't make them in this form factor too much anymore. part number MV50.

First pass was to take a new led and go after it with sandpaper to make it fit. In it goes and the ATR now goes into servo lock (reel servos). Cool, the LED was the problem.

In many ways I am lucky. I work for a University and this University had an electronics lab that was shut down years ago. It also has one of the greatest hoarders of odl parts and a room full of old electronics. After looking for a while I found 2 of the right kind of LED in the parts bins. Asking if there are any more one of the engineers says that he had not seen any of those since the '70s.

With the reel servo working the machine developed another interesting problem. It would go along playing and then go (as I said) skitzo. It would speed up and slow down and when stopped not go back into play.

I cut my teeth on tubes and spent my youth working with TTL. And the ATR is full of TTL. I have an old tektronix TTL test probe. Hook it up to the +5 and ground power and then touch the probe tip to an IC leg and the proble will light up green for a 0 or red for a 1. The color will flash of the logic voltage is out of spec.

That luck part comes into play again as that I decided to trace the play signal logic from the button into the TTL logic. Bingo, just a few gates into the signal path and I find a 7410 that has an output that is out of spec. Digging in my TTL box I find a 7410 and put it in. Now the transport works fine.

More in the next post.....

 

[evm1024]

Heads and audio

With the #2 ATR-100 transport working I need to get a set of 1/4" heads. The machine has a set of 1/2" 4 track heads. Humm, the track spacing and so on is almost the same between 4 track 1/2" and 2 track 1/4". At least good enough for testing. And the 1/2' baseplate guides can be converted to 1/4" just by moving a spacer.

The spacer gets moved, the head connector is reversed to put the lower 2 tracks into the 2 channel electronics (phase and L-R swap but this is testing) and the deck is now playing back Autobahn from a scratch 1/4" tape.

Time to save up some money and get one of my 4 track baseplates converted to 1/4" with some nice heads....

Here is a photo of #2.

--Ethan

 

[sweetbeats]

Wow.

Phenomenal work, Ethan.

An inspirational reminder why I like my MM-1000 because having a "skitzo" deck would make me skitzo and the MM-1000 is heavy on brute force and light on intelligence.

So...why is the control panel on one side of one of your ATR-100's, and on the other side on the other one? Can they be relocated for operator convenience like reversing the hinges on a refrigerator door?

I did not realize that the ATR100 series lacked a pinch roller...how many degrees of wrap does the tape take around the capstan and what is the advantage?

This is really related to the 1/2" deck but what is the advantage of the ATR Services capstan?

Good stuff, Ethan.

 

[evm1024]

Thought you would enjoy seeing that walnut

The ATR control box is removable and can be placed on either side. Moving it from one side to the other is a PITA in terms of all the innerds that are required to be removed to get to the screws that hold the tray. Someday I'll put the pencil tray on one side and the control modules on the same side. Ampexes contribution to lefties....

There are at least 2 versions of capstan for the ATR. One is a machined aluminum capstan with grooves and air channels and the other came with an urethane tire. I think that the aluminum capstan was used for broadcast and the urethane in studio work - but that is a guess.

The ATR services urethane capstan (and I expect MDI Precision to be on par) has very low wow and flutter. They all come with a test report. The capstan in the #1 machine specs at 0.004% wow and 0.016%flutter. The design gives the lowest wow and flutter while not messing with the tape in a pinch roller.

As you can see from the photo there is about 100 or 110 degrees of wrap around the capstan. Urethane has less slip than aluminum capstans during speed changes.

The ATR uses 2 servos. The reel servo is like the Tascam 58 or MS16/atr-60. The supply roller and takeup rollers (ball bearing rollers) are part of a position sensor. Once in servo lock the 2 reel motors supply opposite torque in balance which causes the tape to be tensioned to the correct amount. As the tape is moved by the capstan one roller receives some tape and moves away from the capstan and the other roller has tape taken away and thus moves closer to the capstan. The position sensors detect this and change the reel motors torques to keep the tape tension constant. Like I said this is just like the Tascam decks with the "omega" tape path.

The second servo drives the capstan at a constant rate. Power is applied to the capstan motor which turns an optical shaft encoder on the inside end of the capstan. This shaft encoder (also called a tack) generated a signal whof frequency is directly related to the tape speed. The Tach freq is compared to a quartz osc generated frequency and the difference becomes the control signal for the capstan motor. Capstan too slow and the control signal speeds the motor up, too fast and it slows the motor down.

No pinch roller needed. THe capstan servo moves the tape and very low wow and flutter, the reel servos keep the tape at a constant tension. God help you if your servos get out of whack. Those motors are powerful!

A bit long winded but there you have it in a nutshell.

--Ethan

PS The capstan is on the right and the tape position counter roller (the tach roller) is on the left. The headstack cover is removed in this photo.

 

[sweetbeats]

Wild...just absolutely wild...and those flutter specs are just insane.

AND...I'm not sure how well appreciated it may be just how impressive it is that they (Ampex) designed this system (not sure how unique it is but the particulars are certainly specific to Ampex) of two independent but intertwined servo systems that function THAT WELL together as to elicit those results, and this is on mid-70's gear...AND IT STILL WORKS TODAY!

Ethan, are those reel motors the same for the 1/4"~1/2" transport as is on the 2" ATR-124?