Akai 4000DB / 4000DS

[monaro0162]

Edit below

[monaro0162]

Hi all,

I've recently aquired an AKAI 4000DB in excellent condition to go with the 4000DS that I finished a restoration on in December 2016. This one has very little head wear, and I think has never been opened up before, as when I undid the facia screws, all of them gave a sharp click when they loosened suggesting that they haven't been undone before. Inside it's immaculate with very little dust. It plays perfectly although the clutch mechanisms are squeaky and it has weak rewind.

Assuming its as it left the factory, I am going to use this one as a 'reference' unit to compare to the 4000DS that I have, and another project 4000DB that I have that I am working on and restoring at the moment.

The project unit was working but sounded dull and had a poor frequency response, so I did some recording tests on it by sending frequency sweeps into the unit then playing back the recording into PC HD audio card and analysing the results using Audacity software.

The Audacity software can create graphs of the frequency response and immediately the problem with the dull sounding project unit was shown as a big flat spot around 7khz. For reference I did the same test on the reference DB and it showed a much smoother response with better top end as well.

When I looked at the heads in the project unit I found that the baseplate was fitted with a E4-200/R4-150/P4-150 set (Erase-Rec-Play) whereas the reference unit has a E4-200/P4-154/P4-150 set suggesting that the R4-150 record head is incorrect for a 4000DB. The service manual quotes P4-154 for the record head matching up with what is fitted in the reference DB. Further investigation reveals that the R4-150 was used in the 4000DS MK11 and it is quoted as having an 8 micron gap head compared to the 1 micron gap quoted for the P4-154. So I wondered if the project unit has had the head changed in the past for some reason and the wrong type has been fitted? Only one way to find out - replace the basplate with one with the correct heads. So I removed a baseplate with a set of heads with the correct numbers from another 4000DS that I have. The result - a huge improvement in the sound quality and the flat spot has gone.

I did a few more things including setting the head heights and then took a look at the bias adjustments. The red tamper seal was broken and the screws had been moved indicating that someone has had a go at adjusting the bias in the past. So I reset these whist doing Audacity frequency sweeps and the unit now sounds lovely just as good as the reference DB.

Attached are some photos of the units and the results from Audacity, I hope this is of interest to the community here.

Does any body know the relase dates and the manufacturers recommended retail price (MRRP - remember that!) of the 4000 series. All I have so far is that the 4000Ds was introduced in 1972 @ ??. When was the 4000DB and the MK11 released and does anybody know the prices?

[dave walsh]

"Sounds" like someone thought they could make the wrong heads work monaro or didn't care as long as it seemed to be repaired. Never heard of anything like this before but good write up and great photos. I haven't got any info that would help except that my DB was purchased [from Comet in Rochdale] around 1974 perhaps.

I note that there was a recent posting about cam replacements. This was a constant issue here at one time and a Forum member made some up. Not as easy a problem to solve as it
first appeared!


Dave W

[monaro0162]

Hello Dave,

Thanks, yes I have seen the recent post about the cams. The 4th (and last!) DB that I have has good cams, but requires quite a bit of work elsewhere, so I may dismantle this unit and use it for spares.

For clarification, the graphs posted were all done with dolby off, I haven't got around to doing the same tests with Dolby on.

From what I can gather, dolby should reduce noise or hiss. But from my testing so far (using my ears) it doesn't seem to make a lot of difference, but cuts the top end slightly. May be its better suited for classical music or voice rather than rock (or may be at my age its my hearing!) It will be interesting to see via audacity what effect dolby has.

[TIMTAPE]

Recording frequency response tests should be done at -20db on the VU meter, as per the Service Manual. At higher levels the highs tend to overload.

Dolby only works well when the unit's overall response is quite accurate. The type of music or voice recorded is irrelevent.

Different tape types require different biases.

[dave walsh]

I bought one of the cams monaro, just in case and still have it [somewhere].
Sadly, I think the chap working on that problem died. Things went a bit quiet after that but hopefully the threads are still around. I hope so as there was a big "tidy up" at one point. The UKVRRR site is such a unique and invaluable archive for the future it needs to be preserved.

Dave W

[PaulR]

Before the 4000DS was the 4000D. I bought a 4000DS new in 1972 for, I think, £71. I still have the receipt somewhere. I was always disappointed with the rewind tension. The tape was left very loose on the spool which meant that it slipped when it was played. I took the original one back to the shop as I thought there was a fault but the replacement was exactly the same.

It is now back in its original box in the loft!

[monaro0162]

Here's the suspect head set from the project 4000DB. The record head is clearly labelled R4-150, which was successfully replaced with a headset from a 4000DS, with a P4-154 record head.

Strange thing is the R4-150 is listed for the MK11, yet the MK11 has a different cable connector. This seems to be a mix of the two, it doesn't look like it's been altered by anyone as the wiring & soldering to the heads looks original. Who knows! Maybe qa were asleep when they stamped the baseplate it left the factory like that and has remained that way ever since. Could explain that evidence of fiddling around the bias screws. Poor thing, it was quite battered when I got it, may from someones frustration that it never worked properly!

Originally Posted by TIMTAPE

Recording frequency response tests should be done at -20db on the VU meter, as per the Service Manual. At higher levels the highs tend to overload.
Dolby only works well when the unit's overall response is quite accurate. The type of music or voice recorded is irrelevent.
Different tape types require different biases.

Thanks for the info. I'll re-run the tests using spot frequencies rather than the 0-96khz sweep over 30 seconds that are shown in the graphs. My tests were done with the levels set higher at 0vu at the start with the needle eventually falling to zero once past about 20khz.

So re frequency response in general and dolby, the idea is to move the record levels down to just above the noise floor, then do the assesment there?

I've noticed that the "Equaliser" switch boosts the top end frequencies when playing back, the "tape selector" switch has the same effect but to a lesser extent. Interested to know how each one works. Does the equaliser switch alter the recording bias? I'm assuming the the Tape selector switch just applies (or removes) a filter on playback?

I'm doing all of this using a PC with a 192khz/24bit sound card and the audacity software. In between is my Hi-Fi amp and a mixer, but I'm making sure that they are removed from the signal paths and the audio connections are direct between the AKAI and the PC sound card before making the test recordings.

It would be nice to have all of the old school test kit, and do it exactly as akai instructed, but I don't have the time, space and money to do it that way!

[TIMTAPE]

Originally Posted by monaro0162

Thanks for the info. I'll re-run the tests using spot frequencies rather than the 0-96khz sweep over 30 seconds that are shown in the graphs. My tests were done with the levels set higher at 0vu at the start with the needle eventually falling to zero once past about 20khz.

So re frequency response in general and dolby, the idea is to move the record levels down to just above the noise floor, then do the assesment there?

No not just above the noise floor but 20db below 0 on the meter. The treble is already boosted at the record stage so the tape itself saturates in the highs before lower tones, probably contributing the highs drooping in your tests.

I've noticed that the "Equaliser" switch boosts the top end frequencies when playing back,

Also it changes things in record. As stated on the switch you should set the switch to the tape speed you are using.

the "tape selector" switch has the same effect but to a lesser extent.

Again the tape selector refers to the tape. At that time there were different types of tape broadly described as Low noise and Wide Range. The manual probably lists various examples.

Does the equaliser switch alter the recording bias?

No it affects the equalisation at both record and playback stages.

I'm assuming the the Tape selector switch just applies (or removes) a filter on playback?

No, as above it adjusts the machine to two different types of tape. That may include a change of bias. The switchings were designed to optimise the machine to different tape types including taking advantage of more modern tape types.

On top of that, Dolby NR - maybe we shouldnt even go there- was very fussy about these things and exaggerated even slight misalignments. Amongst professionals Dolby developed a good reputation because they used great machines and had the technical personnel who knew how to set them up precisely for Dolby. Amongst amateurs, Dolby developed - quite deservedly in a way - the opposite reputation. Unreliable, especially on cassette but also not exactly a walk in the park with open reel.

[monaro0162]

TIMTAPE thank you very much for the information, most interesting.

I shall attempt to create a repeatable test routine on the PC, based on the information you have provided, and some more research into tape biasing, dolby noise reduction and tape equalisation.

I'm not intending to replicate the calibration process specified by AKAI, and prove that a machine is working to specification, as that would require the use of old school test equipment, that I don't have.

I'm really just looking at a way of quantifying the existing performance of a machine, as it was received, then making adjustments and making sure that I am seeing step change improvements in performance. Then I will listen to music on the unit, and compare it to the DS and DB 'reference' machines that I have to make sure that the sound, sounds better!

Another issue that needs to be taken into consideration is the amount of background noise (or electrical noise) that the machines produce, when in idle mode and in recording mode, with no audio fed in to the machine. I've done noise traces with the tape selector switched between tape and source, and with the mic and line level inputs set to minimum and maximum. The traces show a number of issues - low frequency 50/150hz mains hum, higher frequency hiss in the circuits (without the tape running) and some very noisy crackling rumbling and waterfall noises if the transistors of LD3141 chips have gone noisy or are failing. I've had quite a bit of success fixing these problems on the DS that I have, by replacing all of the transistors and caps, and by improving the signal earthing points.

So I have lots of work ahead

[dave walsh]

Between yourself and Tim this is a very good info pack on the 4000. My only major problem has been drop out and crackles via the Track Change switch. I could never detect much difference with the equaliser so either it's out of spec or I've got "cloth ears".
I've used Servosil on the switch [after getting the cover off with a mini allan key] and various bodges to keep the channels clear but it probaly needs complete disassembling/re-soldering [plus courage in my case]. Primarily I've still got lots of material to transfer so I'm using the "leave well enough alone" [Popeye's song] principle at the moment. I don't record on Reel to Reel currently but may go back to it. I've also got a Tanberg 3300X as back up. That's pretty good and compatible with my tapes despite me never liking the use of a joystick that much.

Dave

[dave walsh]

Apart from Tim's reference to tests at the noise floor, I've just realised that you did the "Re A Lovely Akai 4000DS" thread on 10/12/16. Very detailed and including an auto-stop mod to enable the 'motor noise' to be switched off but still retaining the abilty to listen to source material-something that I'd not seen described before

These Akai's used to get dismissed at one time on an "only one motor and jumped up design basis" but especially in the 70's, the 4000 brought a number of professional features in one clever package to users who would never be able to afford the top end studio gear. They were great then and still are, hence their success. It was a bit like the Alan Sugar story-not my favourite personality but his home computers were a tenth of what you would have to pay for a basic IBM machine at the time and flew out of the door. People were really excited by both products.

Dave

[monaro0162]

Originally Posted by dave walsh

Primarily I've still got lots of material to transfer so I'm using the "leave well enough alone" [Popeye's song] principle at the moment. I don't record on Reel to Reel currently but may go back to it. I've also got a Tanberg 3300X Dave

I have many reasons for having an interest in these machines: My Dad had one from new, which I used to use a lot, I like vintage electronics and hi-fi, and I am a big music lover. I am a member of www.dimeadozen.com* where historical live rock band music recordings are shared using lossless FLAC files and bittorrent. Often these recordings were made back in the day onto tape or cassette from radio broadcasts, and now years later, they are being transferred back from tape into digital format, in lossless form, for all to enjoy. Like going back in a time machine! It so happens that the first DB I bought came with a couple of Eric Clapton "off the air" reel recordings which I'm keen to transfer to digital, after the machine has been setup as best as possible.

*dimeadozen it not a pirate site. Only unofficial live radio broadcasts, or diy recordings, or studio outtakes are shared, with no objection from the artists. No official releases are shared there. For those you can use Itunes or similar.

[monaro0162]

Originally Posted by dave walsh

These Akai's used to get dismissed at one time on an "only one motor and jumped up design basis" but especially in the 70's, the 4000 brought a number of professional features in one clever package to users who would never be able to afford the top end studio gear. They were great then and still are, hence their success. It was a bit like the Alan Sugar story-not my favourite personality but his home computers were a tenth of what you would have to pay for a basic IBM machine at the time and flew out of the door. People were really excited by both products.
Dave

I agree. These 4000DS's are often dismissed by 'professional standard' recorder owners as being mickey mouse, conversely some DS owners describe the DS as a 'professional/semi-professional' recorder, but in reality they are a domestic recorder built to a budget, that actually offers pretty good performance

[Ted Kendall]

I bought a 400DS as my first stereo machine, and was pleased enough to have it for £65 from Lasky's and tote it home on the bus. I did get a lot of fun out of using it but changed to a Sony TC-377 a couple of years later, and this was undoubtedly a better machine.

The weakest point of the Akai deck design is the use of a capstan sleeve to change speed, closely followed by the temperamental and complicated clutches, both of which affect speed stability. Looking like a left-handed copy of the Ampex 600 series didn't help either. However, on a fine day with the wind behind it, a 4000 series can give good results, and it was certainly stonking good value at the time!

[monaro0162]

That's interesting. I have a low hours use TC366 donated to me by a friend, who's father owned it from new. Hardly used. My friend modified it years ago by re-purposing the microphone potentiometers to alter the bias, but today he can't remember how he did it!

On the weaknesses of the Akai, I would have though the opposite. As the capstan and pinch wheel are bigger relative to the ones fitted on the Sony, shouldn't they provide more tape 'grip' and suffer from less w&f, if the pinch wheel rubber gets damaged? Or may be I'm thinking wrong. I like the simplicity of changing the running speed by just fitting the bigger capstain, rather than relying on a mechanism to do it as on the Sony?

I haven't run the Sony much yet initial checks reveal that it plays, but randomly slows down after a while. Probably needs a new belt. I don't like the big rotational knob to engage play on the Sony, as when moving it into the play position, there's no confirmational 'clunk' that its in the right place, as you get with the DS. I had a bit of trouble with the Sony when I got it, it wouldn't stay in the play position and kept flipping back. Only thing I've done so far is apply a bit of lube and it now seems OK. May be they were just made like that?

[Ted Kendall]

Whatever is wrong with a particular example, I don't think there is any real doubt that the 366/377 deck is a better device than the Akai in performance terms, The runout caused by any capstan sleeve which is not a snug fit in the shaft is enough to cause audible flutter, and anything tighter is likely to score the shaft as soon as any oxide is deposited on it - cleaning the shaft on my machine was a constant chore. The dimensions of the Sony capstan and pinch wheel are approximately those used in the Revox A and B 77 series, so there is nothing inherently wrong with them.

I didn't find the Sony rotary control awkward once I'd had the machine for a couple of days, and the three speed idler transmission, whilst utterly conventional, is effective and reliable. Furthermore, Sony took a couple of features from Telefunken domestic designs which lift the performance of the deck above the norm for the price. One is the use of a mechanical servo to provide constant back-tension and the other is the use of a scrape roller by the record head. Both these features are commonplace in professional machinery and contribute to the remarkably clean sound a 377 produces when in good order.

I repeat - the 4000s were great budget decks and I loved mine, but the 377 was a better tool. It served me well until I blew a term's grant (oh, yesterday, leave me alone!) on an A77.

[monaro0162]

Originally Posted by Ted Kendall

The runout caused by any capstan sleeve which is not a snug fit in the shaft is enough to cause audible flutter, and anything tighter is likely to score the shaft as soon as any oxide is deposited on it - cleaning the shaft on my machine was a constant chore. The dimensions of the Sony capstan and pinch wheel are approximately those used in the Revox A and B 77 series, so there is nothing inherently wrong with them.

Oh I understand now. Provided the capstan is a perfect fit on the shaft, then all is good. If there is any play between the two, off-centre movement of the capstan, as it rotates, will cause a variation in the pressure applied on the pinch wheel ... causing wow and flutter. I have to say that I haven't noticed any wow and flutter on the three AKAI machines that I have running, and there is no perceptible running speed difference between the three machines either. Generally I tend to prefer minimalist engineering solutions, so the simple AKAI speed change method is fine with me

Once I've finished the DB, I'll get the Sony running, and provided I can fix its tape slowing to a halt problem, I'll do some comparative performance tests between the two machines. Interested to see if it is any better than the AKAI. As well as measuring the frequency response, I need to find a way (using a PC) of measuring wow and flutter. As far as I know Audacity doesn't do it, as it's primarily software for creating very good digital recordings, rather than an audio analysis tool.

[Ted Kendall]

Actually it is the runout itself which causes the flutter, as it changes the instantaneous diameter of the capstan.

As an aside, similar debates have rumbled on for years between Ampex and Studer devotees...

[monaro0162]

Originally Posted by Ted Kendall

Actually it is the runout itself which causes the flutter, as it changes the instantaneous diameter of the capstan.

Got it.

Out of interest, are the innards (and circuitry) on the 366 similar or the same as the 377? A quick check on the net suggests that the 377 runs up to 30khz, but at what +/- db I don't know.