DCC 170 problem after belt replacement (solved?!?!?)

Dear All,

Upfront, as always a big thanks for your support and feedback. I really appreciate this community!

After successfully replacing the belt on my DCC 175 I wanted to have a DCC 170 as well :slight_smile:.

Once I managed to buy a DCC 170, my tests revealed that the motor is humming but no display is shown. The solution in most cases can be found in replacing the belt like I did for the DCC 175. So all went pretty okay until I added the new belt…

On this DCC 170 I removed the two PCB boards to finally reach the motor, belt and gears section. The old belt was pretty warn-out so I removed it. I first cleaned all the belt wheels (three copper/metal wheels and one black plastic one) with IPA to remove remaining rubber/grease. Then I placed the new belt which fitted fine.

So far so good… but then during manual turn tests, to check the belt’s position by switching the black lever a few times, I noticed that it works fine until the black gear’s area (highlighted by red circle) reaches the black lever position (green arrow). Then the belt starts slipping and the blue circled switch lever will not change its position. For other two black gear wheel areas/positions this is not a problem and the blue circled switch lever nicely change its position, except for the red circled area.

To overcome this I tried some grease/silicon spray on each of the wheels and gears, but still no luck. the belt seems to be okay, not to wide/tight. So any idea what causes this behavior? Seems like once the black gearwheel turns to the red circled area angle, it delivers some resistance with the black lever and as a results the belt starts slipping. At this point it is not blocked as, with some manual help, it can complete its turn resulting in the blue circled lever to change its position. Please see picture (taken from internet to clarify) below to explain the issue according colored circles and arrow.

I hope I explained it well enough for you guys to understand as it is difficult to write down the exact issue.

John (Lee Fang)

An update …The problem is somewhat solved by replacing a new 0,6mm belt instead of a new 0,8mm belt. Now the gearwheel can make makes nice turns when the black lever is “set/raised”, meaning the blue circled lever switch can reach all three positions. BUT… the DCC 170 now plays, but hampers a lot during playback. A service mode test shows the following:

Service Test Format: 0-000-00000000 translated in W-XYZ-UUUUUUUU to explain the issue
The Z value (AUX Data error) constantly changes from values between 0 to 6, while W value (unreliable data) regularly changes values between 0-F. If this W value changes from 0 to any other values between 1 and F, then the sound is dropped. The remaining X,Y and U values never change during playback.

Any idea what this means? So almost constantly during playback I see “AUX data error” and quite often “Unreliable Data”. The “Speed Adjust” shows a value of 0330 on my DCC 170 while it shows a value of 032D on my perfectly playing DCC 175.

So I am open to any suggestion :slight_smile: and as always thanks in advance for your any feedback or thoughts.


1 Like

This could be a speed or head issue. Is the belt 64 mm?

Hi Dr DCC (mr Ralf :slight_smile:),

First of all thanks for your reply!

The belt is indeed 0,64mm and seems to work fine (based on manual turn checks) compared to the 0,8mm belt. But today I also tested with an analog cassette which sounds like a fluctuating high pitch/speed. This doesn’t sound normal at all :), even adults voices sounds like singing children :slight_smile:. For comparison, on my DCC 175 the analog tapes also sounds somewhat “higher” but at a stable pitch/speed.

I can order a new 0,8mm belt or do you thing that speed might be the issue. Not sure how to change the speed (in the speed adjust mode) without having the analog test tape (as mentioned in the service manual).

Thanks all in advance for any of your inputs :)!!


You might just want to try with a custom 3000Hz tape.


1 Like

Some are commercially available, I made one myself with an okayish tapedeck and that was good enough.

Dear Dr DCC and Max,

Thanks for your feedback and I did some research on the 3000Hz tape.
As I don’t understand the concept of changing speed in combination with a 3000Hz tape, i have some questions (Sorry :slight_smile:).

  • This test tape does it need to be an analog cassette or DCC (i guess analog from what I have read)?
  • Can i record a 3000Hz signal from YouTube or generate it myself via sound oscillator software?
  • How does the speed adjustment work? So set DCC in Speed adjust mode and play the 3000Hz sound and then what … will it auto adjust or do I need to turn a potmeter, anything else …???

Sorry for the questions (as internet did not reveal enough details to make some sense out of this).
Thanks in advance for your replies!!

Learning a lot mainly through this forum and by breaking/fixing DCC recorders :slight_smile:


The only way is to either buy a 3000 Hz tape or create one.
Creating one is never really a good option, unless you have a fully calibrated deck.

I believe the steps are in the service manual on how to adjust the speed.


On a side note… I’ve thought for many years (since the 1980s) on how one could calibrate the speed of a cassette deck. With the current state of generally available software and hardware, this should not be that hard.

I would probably start by making a tape of an exact length, e.g. 18 3/4 inches (47.625cm) which is supposed to be exactly 10 seconds duration if a recorder runs at the correct speed. I have a tape splicing kit that would let me do that.

I would put the tape with that exact length in a cassette (of decent mechanical quality) surrounded by lead-in / lead-out tape, and then record a 1kHz tone from a precise oscillator (this is easy to accomplish with a computer running software such as Audacity).

Then I would play the tape and count the number of pulses on the tape in some way. I’m not sure how I would do that but I imagine there is software that can do it, or I can write a program to analyze a WAV file recording and generate a number of times where the samples went from negative to positive or something.

So now I’d have a number that indicates how many audio cycles are on a piece of tape that should be exactly 10 seconds. The actual tape speed can be calculated based on the number of cycles and the knowledge that the desired number of cycles should be 10,000 for a 10 second recording. Obviously because of wow and flutter it’s a good idea to record the tape several times and get an average of the counts. The standard deviation of the counts might give an idea of how much wow and flutter is in the recorder, but that’s a different story.

So for example if the recorder recorded 10,100 pulses on a 10 second piece to tape and you know the pulses were recorded at exactly 1000 Hz, you know that the recorder is running at (10,000 / 10100) times the required speed. If the recorder recorded 9000 pulses, the recorder must be running at (10,000 / 9000) times the required speed.

Now you can change the frequency of the oscillator to compensate for the speed of the recorder. For example if the count was 10,100, you set the oscillator to (10,000 / 10,100) * 1000Hz = 990Hz. If the count was 9,000, you would set the oscillator to (10,000 / 9000) * 1000Hz = 1111Hz. Then re-record the tape with the new frequency, and re-count the total number of pulses on the 10 second piece of tape on playback. If the recorder is still running at the same speed as before with not too much wow & flutter (regardless of how much the speed adjustment of the recorder is “off”), you should get pretty much exactly 10,000 cycles recorded on the 10 second tape. At that point you have a piece of tape that has a recording that (if played at the right speed) is exactly 1kHz!

Without adjusting the speed of the recorder or the frequency of the oscillator, you can now record an entire tape, and you’ll end up with a tape that’s essentially calibrated to 1kHz. You can use the tape with any recorder (including the one on which is was recorded) and use a 1kHz oscillator to adjust the recorder speed to the exact right setting. You won’t even need complicated equipment or complicated steps or complicated math from then on: Just put a 1kHz oscillator on a speaker and the tape player on another speaker (or put both on the same speaker, it doesn’t matter) and listen to how the two 1kHz waves interfere with each other.

If the speed is off, you will hear the two frequencies “beating” each other. By adjusting the speed in the recorder until the “beating” comes to a stop, you can calibrate your player to an exact 4.7625 cm/s speed. Again, you will always have some wow & flutter so it’s probably not possible to totally make the “beating” stop but I’m sure you can get very close, especially with a high quality cassette deck.


I used a instrument tuning software, don’t know how it is called.

Interesting Jaq.

Analogous to a metronome, you have a fixed number of pendulum movements in time. Just like the 1 kHz sine wave that fits on a piece of tape of a fixed length. Then you play it back and adjust the speed by synchronizing the pitch by ear.

It’s like tuning your guitar.

What I don’t understand yet is the number of pulses you describe.

Just buy one from Mihokm and be on the safe side.



Seller is a well known specialist in casettes and analogue technique.


Dear Ralf, Jac, Max, Jos and Chorazy,

First of all sorry for my late reply. second, big thanks for your replies!

Let me digest this info so I can makes sense out of this and then decide what to do (make my own tape or buy one). I really appreciate you sharing your knowledge on this!


This. That is exactly what I have with some of my portables too. I tried the same thing that you have done.

I found out that the lever in the blue circle is responsible for the switching of the capstan pinch roller from forward play into reverse play and vice-versa.
In fact, it moves this whole area from one side to the other:

One of the 2 curves metal poles in the circles get pushed out by the grey metal blobs when that direction is selected (FORW-REV) and the metal bar at the bottom (between the red lines) goes with it.
I find that that movement from left to right (or vv. ) has to overcome a LOT of friction, and that causes the black gear at the far right in your picture to ‘jam’ and that then causes a slipping of the belt at the motor while the motor is still turning.

I have ordered tighter belts, 58mm by 0.6mm instead of the ususal 64mm or 67mm by 0.6mm, and that helps somewhat but the friction is still too much so the belt still slips at the motor.

I tried to grease up that entire red area, but it still won’t work. I’m stuck at the moment.

1 Like

Hi pvdm,

That was also my finding. The movement of the metal with the two curves causes to much friction resulting in a slipping belt (only in the specific lever switch/black lever position). The 0,64mm belt is somewhat tighter and now that problem seems to be solved. I also greased the specific area you showed in picture above. Another finding: the 0,8mm belt doesn’t seem to be a good fit (compared to DCC 175) as I had to remove a small part of the plastic (with the iron bolt) to get the belt properly around the smallest copper wheel. This was needed to prevent the black plastic to get in contact with the rubber belt on the smallest wheel. But in the end the 0,8mm belt didn’t work as mentioned above.

I still have problems and these might be related to this smaller 0,6mm belt or due to incorrect tape speed. Still not clear why it intermittently plays DCC cassettes (lot of dropouts) and why analog cassettes sound weirder on this DCC 170 than on my DCC 730 or DCC 175.

So pvdm, I hope you can tackle the issue!


We have done nearly 150 portables. All the different models, and without exception they work with a 0.8mm belt.

I have seen however these problems with a 1mm belt.
Are you sure about your belt size?

1 Like

I will see if I have a 0,8mm belt lying around and try that.

Dear Dr Dcc,

I had to remove a minimal/small piece of the plastic near the small copper wheel (see screenshot - red eclipse) because the belt (0,8mm according my caliper :slight_smile:) was really to thick! But in the end the mechanism that @pvdm showed in picture above resulted in so much friction that the belt slipped and the plastic gearwheel (and therefore switch lever) never reached the right position. The 0.6mm belt seemed to solve this slipping, but now I have other problems (intermittent playing, so this could still be the belt or motor speed).


Maybe some grease ended up on the 0.8mm belt. I would try to wipe it with Isopropanol or similar.

1 Like

That. I need to find out why that mechanism that switches the pinchrollers is so hard to operate.

Oh, and have you tried an analog cassette, preferably a speed calibration cassette, or one with music containing long notes, just to be able to hear what the mechanism is doing?

1 Like