34B hi freq loss on record

[evm1024]

I'm working on calibrating a 34B and I've run into a strange problem. The history of the machine is not known.

On playback I get normal freq response from 31.5 Hz up to 20 kHz. Flat +- 3 dB. Good High frequency response. Listening tests sound fine.

Or record (250 nW/m to fresh 456) I see a distinct high frequency rolloff to about 15 dB down at 20 kHz. Adjusting the record EQ has very little effect. The rolloff starts at around 4 kHz.

Looking at the guides and lifters I see very little wear. The record and playback heads look alike in terms of wear. They do not have the mirror like finish that I would expect but rather a rougher surface. On first look they look like they have low hours.

At this point 2 thoughts come to mind. First is that the EQ switching logic has failed. The second is that the guides and lifters were replaced and that the heads were relapped and perhaps not done that well.

I'm not experienced in looking at heads so here is a photo if you would care to lend your experience and comments.


--Ethan

 

[Tim Gillett]

evm,
It could be a number of things, singly or added together.
You are right to inspect the heads physically. Hard to tell this from the picture. They look OK but.... The tape path surface should be mirror like.
The record head's vertical gap should be almost invisible at normal viewing distance. if there's a strong black vertical line showing, the head's gone, or at least the track segment which has the line is gone. Both heads should be physically aligned so there's intimate contact at all head gaps.
The record head's azimuth must be lined up to the play head's azimuth (the play head is probably not too far off if it replays a standard test tape OK).

High frequency response is sensitive to record bias level. This must be checked after the other things are checked. (In short for all of this get the service manual and follow it to the letter!) When adjusting record parameters, especially frequency response, use a lower recording level, such as -10 VU or even lower. A modern tape like 456 should be able to handle tests at 0VU. The higher then tape speed the less critical. But there's no point in pushing your luck. Be conservative. The reason is that the highs overload before the lows and mids and then they just compress down. If you keep pushing the highs even further you actually get an inversion. The more you boost the highs, the less you get off the tape. Best to check frequency response at a recording level well below 0VU.
Record heads as they wear change their inductance. That means that depending on the type of oscillator feeding it, the bias level that was appropriate when the head was new may not be the same as the bias level required when some of the metal has worn away. The tendency is for the record head to gradually become overbiased, which wipes out the highs.
If you dont know this you might conclude a head is useless simply because the highs are down. It may have just need a bias tweak.
I suppose it's possible the record eq is faulty but I'd be checking these other head and bias level issues first. They're the more likely culprits. I'm not that familiar with this model. Maybe there are other issues specifically related to the 34B that others on the thread can help you with? As in the manual, record eq is done after bias adjustment as a sort of fine tuning.
Hope this helps,
Tim

 

[SteveMac]

Sounds like a problem similar to what I was having. But I was able to fix it by doing the normal playback/record calibration and azimuth adjustment and it was only off by about 3 or 4 db.

 

[evm1024]

2-11-11 Frequency Response overall

Tape path alignment is fine. Azimuth was scoped. All cailbration steps prior to 2-11-11 Frequency Response Overall were fine. Bias set to 4 dB over.

Sorry about the poor quality of the photo. I have seen a worn out head where the gap started to widen. Don't see that here.

I'm thinking about that roughness and asking myself what was done to the heads. I might try polishing them to see what will happen. (per noretronics handbook)

In any case I'll restart calibration from step one just to be sure that nothing was missed.

 

[cjacek]

I think the head photos are very good. Speaking from experience of having bought an unused 34B, I can see the curvature of the heads being similar to mine. The heads, tape lifters etc ... look like they never or very rarely passed tape over them. The only thing I'm worried about is the lack of a mirror like finish, as was already suggested. Could it be some oxidation starting to show .. or a chemical being used ? I don't know but there are specs of some greenish / greyish substance there ...

 

[Dr ZEE]

if i'm not mistaken, those heads are not shiny as they are made.
as for the photo:
the areas of the heads (that looks on my screen as rather blue-ish) are not polished-like (unlike some many other heads, which are pretty solid-shiny all around). what ever it is made of - I have no clue, but that's the way it is. If you look at the heads with natural light - the color is rather grey (or silver-grey, but it's flat, not glossy). If you point some light on the heads, those areas can reflect the light and look pretty blue from some angle.
also, these heads shape (when new) is somewhat pointy and very narrow area (that is in contact with the tape) may be somewhat polished and so look (or say - be) polished and shiny (maybe somewhat 1-1.5 mm wide).
The inclosure of the head (play and record) (I am not sure if "enclosure" is what it's called ... , ) is made of some different metal and is pretty shiny, also I don't think that the condition of the enclosure has any effect on head's performance. The erase head got no shiny metal at all - poor deprived baby .


I'd say, the heads look very good... from what I see.

/later

 

[evm1024]

More data

It looks like the heads had some type of residue on them. I was able to do a little better cleaning and I have to agree that they look near new. The playback and record heads look much the same. Here is a better picture.

The problem is in the electronics. I have not solved it yet but at least I know where it is. I'll spend some time looking at the schematics and get it solved perhaps this weekend.

It has become more interesting. Here are some of the puzzle pieces:

1) playback of a MRL tape yields a flat frequence response. This tells me that the playback head and electronics are OK and could be calibated.

Then record a 20 kHz signal onto 456 at -3 VU and:

2) in playback mode playback of that tape through the playback head gives -10 VU ( 7 dB loss, bummer!)

3) in playback mode playing back the same tape through the sync heads with the record function switches off gives -10 dB (the same 7 dB loss)

4) in playback mode playing back the same tape through the sync heads with the record function switch on (channel by channel) gives no signal from the tape for that channel. Very odd.

This tells me that the record relay is energizing in reproduction mode which should not happpen. That is not the cause of the problem I am seeing but it does indicate that there is a problem with the logic. And I susect that that is the cause of more than one problem.

Like I said I'll have to look over the schematic over the weekend and try a few measurments to see if I can find the "logic" in the failures I'm seeing.

Get this fixed and I would consider this a major eBay score inthe form of a very low hours 34B.

The heads look very shiny now with a mirror like finish.

 

[evm1024]

Feature not a bug

Turns out that case 4 is a feature. In reproduce mode depressing the function select will cause the input source to be selected rather than the head.

Back to thinking out what could cause this HF loss.

Hope you all don't mind me thinking out loud. I do value your comments.

 

[Dr ZEE]

Hope you all don't mind me thinking out loud. I do value your comments.

My comment is: I'm reading and looking forward for your final report and conclusion, ... This is all I'm capable of
.. and thus, obviously, I don't mind you posting your final conclusion and any details that led you to it.

 

[SteveMac]

My comment is: I'm reading and looking forward for your final report and conclusion, ... This is all I'm capable of
.. and thus, obviously, I don't mind you posting your final conclusion and any details that led you to it.

Same here.

 

[evm1024]

And the solution is.........

I spent some time over the weekend looking at this problem (Sorry to say I neglected my wife. Oh what analog causes us to stoop to, how dumb)

I took the channel cards out and hooked up some test wires and found that the logic was indeed doing what it should do. THe opamps were rolling on more HF as the freq was going up. Still no solution.

Then thinking cool I have a tascam 32 and they have the same channel card so I moved one from the 34 to the 32 and it worked fine. Let me say that again the card was working fine.

SO what does that leave us? The heads. Hm....

As you may recall I speculated about the age of the heads and guides. Are they original or replacement/relapped. Looking at them for the last week I would say that they are original and that this is a low hours machine.

I did note that the heads were not mirror smooth as we would expect and thus I thought that I had nothing to lose so I got a new buffing wheel for my mototool and polished the heads. The goal was to restore a mirror finish but not change the profile. I did not use any buffing compound, just the buffing disk going round at 30,000 rpm.

The end results was the restoration of HF recording to within 3 dB at 20 kHz. This was after just a few moments with the buffing wheel. I had cleaned the heads more than once by hand. They are much closer to having a mirror finish. I will buff the heads a bit more as I have time to do a good job.

I'll post a photo when I have them like I would like to see them.

So the answer is that the crud on the heads was thick enough to limit the HF record effeciency but not so thick that the playback EQ adjustment could not compensate for it.

EThan

As a side note any 34b heads in the states? From Tascam? I do see that guy in Japan that sells them for about $80 each.

 

[cjacek]

so I got a new buffing wheel for my mototool and polished the heads.

Ah so! A "poor man's" relap!

The end results was the restoration of HF recording to within 3 dB at 20 kHz.

GREAT!!

As a side note any 34b heads in the states? From Tascam? I do see that guy in Japan that sells them for about $80 each.

That's certainly very inexpensive, especially in comparison to other makes / models. You have a very low use 34B and it may be worth it to get, at the very least, a replacement head stack for the future ...

 

[Beck]

Really nice work, Ethan!

I don't know the price for the 34B heads, but TASCAM Parts should have them. They are probably a bit less than 9500 yen. For comparison, the Japanese site has the heads for the 32 at 6300 yen ($54.00), but in the States they are around $40.00 last time I checked.

The rec and play heads for the 34B are the same part number, 5378301400. (You probably already knew that).

~Tim

 

[The Ghost of FM]

Or record (250 nW/m to fresh 456) I see a distinct high frequency rolloff to about 15 dB down at 20 kHz.


--Ethan

I believe the bulk of the problem lays in the tape you're using.

456 tends to loose highs on a pretty universal basis from what I have personally experienced and read from many other users of it. Switching to Emtec 911 may show a marked improvement in high frequency retention and level.

It's also critical to keep the heads meticulously clean when using 456 as any little bit of oxide build up will reduce your tape pressure and reduce the machines ability to print high frequency information to the tape.

Cheers!

 

[SteveMac]

I believe the bulk of the problem lays in the tape you're using.

456 tends to loose highs on a pretty universal basis from what I have personally experienced and read from many other users of it. Switching to Emtec 911 may show a marked improvement in high frequency retention and level.

It's also critical to keep the heads meticulously clean when using 456 as any little bit of oxide build up will reduce your tape pressure and reduce the machines ability to print high frequency information to the tape.

Cheers!

How's it going GFM?

I know it's been said here somewhere but is EMTECH 911 bias the same as 456. Where do I get it? I'd like to try it out.

 

[jpmorris]

How's it going GFM?

I know it's been said here somewhere but is EMTECH 911 bias the same as 456. Where do I get it? I'd like to try it out.

You can occasionally pick up EMTEC or BASF 911 on ebay. The stuff currently in production is branded RMGI though.

*EDIT*
And it is supposed to be compatible with 456.

 

[Tim Gillett]

Ethan,
You could have saved yourself some trouble!
In your first post you said that the heads didnt have the mirror like surface like you would expect.
I confirmed that they should have a mirror like surface and that if they didnt you should expect HF to be down. Also that you should check these head issues out before trying less likely possibilities like incorrect record eq.
In your later post you told us "the problem is in the electronics"...hmm.

Anyway glad you've sorted it out.
Tim

 

[evm1024]

No trouble at all

Hi Tim G,

It was no trouble at all to be systematic at this. I did learn quite a bit. As you may recall I did note the condition of the heads. And rather than just jump into aggressive head cleaning I elected to manually clean them really good and then to step by step verify each system. Once the logic and signal paths were shown to be working within parameters I went after more aggressive head cleaning. Intending to lap the heads if needed.

Also you might note that I have not listed everything that was done to debug this. More than one reel of tape was used, the tape was tested on another machine to be sure that it could be printed at 20 kHz at 250 nW/m, and the head part numbers were checked to be sure that they were the correct heads and so on. (Hats off to Beck and Ghost)

The way I look at it if someone else can learn something from this thread then I have added to the collective knowledge of the list.

Your points were well taken and reflect a good understanding of analog recording and were appreciated. I'm sure that your last post is not meant to be a "told you so" post

Regards, Ethan

 

[SteveMac]

You can occasionally pick up EMTEC or BASF 911 on ebay. The stuff currently in production is branded RMGI though.

*EDIT*
And it is supposed to be compatible with 456.

Thanks, I'll have to check that out.

 

[cjacek]

I know it's been said here somewhere but is EMTECH 911 bias the same as 456. Where do I get it? I'd like to try it out.

Yes, I believe the Emtec SM911 is bias compatible with 456. You can get the tape here:

http://www.usrecordingmedia.com/opreelrectap.html