Akai 707 Preamp Modification

[untune]

Hi all,

Relatively new here, built a Bo Hansen DI and a few bits of modular synth stuff and I'm looking to tackle something bigger. I was going to make my next project a valve mic preamp but I happened upon this Akai R2R - it's very similar to the M7 and M8, and the various Roberts decks that are often used as monoblock amps, except the valve compliment is quite diferent and consists of 2x 6AU6, 1x 6BQ5, 1x 6AR5 & 1x 6X4. This particular model is seemingly quite rare, perhaps having more in common with the earlier Terecorder designs but having a more unusual tone adjustment setup. I'm not sure if there's a Robers equivalent but the 900 series seems to look very similar. The VU meter is dead but I will replace that if I can source a decent, vintage one (which is proving difficult - if anyone has anything spare please get in touch!)

Anyway - I've not tackled a valve project before and was hoping some of the more experienced amongst us might have some helpful input and guidance. I've managed to get hold of and clean up the schematic, and the first port of call is going to be eliminating the unnecessary recorder-related components from the circuit. Any help in what can be removed and what needs to stay would be grand.

First observations:

• A test showed that, during recording, there is a significant treble boost which took a high cut of around -28db to roughly match the same signal recorded through a Warm Audio WA12. I assume there is an IEC treble boost being applied which also comes through the output jack. I'm not sure where in the circuit this is introduced.
• The tone controls are not active during recording, only playback - however, the switch (bass-flat) does introduce a nice bump somewhere around 100Hz when set to bass. I'd like to keep the tone control option present and perhaps have the option to bypass the pot control.
• Noise and whoosing when adjusting volume. I'll put this down to the fact that I've not yet recapped it, and the valves might need swapping.

From what I can see:

• SW1 can be eliminated, and things can be bridged as if the deck was set to 'record' mode - not entirely sure what else can be bridged though. Similar situation with SW3.
• All tape mechanism electronics can go, power/voltage selector etc can go - 3 way power/amp switch can be replaced with an ON-OFF DPDT with a fuse inline, it'll be operating on UK mains voltage.
• 6AR5 isn't necessary without the tape section? If this is the case I'll need some help figuring out what to take out and how that part of the circuit should look. Oscialltor coil can go, I imagine.
• Not sure what the output impedance is, and the output transformer is unbalanced - would there be any point modifying this for balanced output? Would that affect the speaker out?

I'm hoping to end up with a decent microphone preamp that will most likely be used as a line amp for synthesizers/instruments etc. Goals are:

• Keep the line and hiZ mic inputs in place. Plenty of old mics that will work with this and I have some vintage lo-hiZ transformers from British tape machines that I can utilise with newer mics.
• Eliminate the useless head output, and (if possible) have a direct input for bass guitar.
• Add a transformer balanced mic input on an XLR socket, with polarity switch and pad.
• Speaker output in place of the speaker, if it would be logical to use this as an amp too (present speaker is 8ohm/4W.)

I'm sure there's more but there's a lot in my head at the moment. I'm starting to think the best option is to tear the whole thing down, clean and replace everything that needs it, then build it back up from scratch without the unnecessary parts. I'm more interested, given my lack of experience with valve circuits, how the core preamp circuit might be improved or altered given that the valves are different to those found in the units that are usually modified. As I understand it the pentodes have considerable gain and may be a little too much. Opinions?

I'll attach the diagram and also a modded diagram from a project I managed to find on another forum, from a similar machine with the same valves (can't remember where, sorry) which has been whittled down to a more basic circuit. Might be a good starting point if nothing else.

Any help is vastly appreciated

Cheers!

 

[untune]

And here is the diagram I found elsewhere

 

[gridcurrent]

.

 

[untune]

I can't access that attachment - has it been removed?

 

[ruffrecords]

I cannot access it either - this happens quite a bit lately - might be something to do with the ongoing tweaks to the site.

Regarding the circuit, there is certainly too much gain. The second circuit seems a step in the right direction but has some oddities. I would be tempted to leave the first stage relatively unchanged. The 6AU6 is a good pentode but you will need to select one for lowest noise (I used these in my very first mic pre design and I tested loads of them for noise). The second and third stages could possibly be combined to create a lower gain but much lower distortion output stage by adding some negative feedback between them. The oscillator tube I would leave in place but disconnect all the leads except the heaters. This will ensure the heater load is correct. I would also be tempted to improve the HT decoupling the reduce hum.

Cheers

Ian

 

[pvision]

I've just bought a pair of AKAI M8 preamps and plan to do pretty much the same thing with them

My first question was whether to use the record side or use the playback side as the basis for a preamp? There will be pre-emphasis on record & corresponding de-emphasis on playback and would guess the two functions use the same eq circuit in inverted fashion

I was planning to have transformer in & out, a mic / line / di input on the front panel and a balanced output, possibly with a switchable attenuator to reduce output for a DAW

The M8 power in feeds the transport section. A multipin connector feeds the top (left) preamp from the transport and the there's a US-style two-pin connector from the right channel preamp which plugs into the left channel. I assume this means the preamps use a 110 Volt supply but, then again, assumptions are the mother of all cock-ups so I will get my meter out to check

I haven't located a schematic yet but will post it up when I do

Nick Froome

 

[pvision]

It seems that many of these valve preamps (M7, M78, R2R) are very similar

Page of mods here:
http://lacinato.com/cm/audio/robertsmod

Suggested mods from DIYaudio
http://www.diyaudio.com/forums/tubes-valves/214339-request-suggestions-akai-m8-mono-amp-schematic-validation.html

Rebuild walk-through
http://www.diyaudio.com/forums/tubes-valves/166512-akai-m8-rebuild-up-running.html

Evidently it's based on the Mullard 3-3 Three Watt Amplifier

Schematic
https://groupdiy.com/index.php?action=dlattach;topic=64477.0;attach=46548

Nick Froome

 

[ruffrecords]

I've just bought a pair of AKAI M8 preamps and plan to do pretty much the same thing with them

My first question was whether to use the record side or use the playback side as the basis for a preamp?
Nick Froome

H Nick,

This depends on the particular tape machine. Some use the same electronics for both record and replay and use complicated switching to change EQ and input/output routing. I  tnis case you use the one amplifier and strip out the EQ and switching. Other have separate record and replay electronics. Both are designed to take a low level signal, EQ it and raise it to line level. So you could make a pair of mic pres out of this by again stripping out the EQ.

Cheers

Ian

 

[untune]

Thanks for your input Ian - so when we say relatively unchanged at the first stage, is that to say that all resistor and cap values can be left alone? The first 6AU6 has a 1uF electrolytic between screen and cathode whereas the modified diagram has a 0.1uF. How would changes here affect the first stage? How about the option of a DI for bass guitar - would this be inserted into the first stage or would it not work particularly well?

I will have to somehow audition a few valves for noise when it gets to that stage, or hopefully there'll be someone who might provide a service for selecting a good pair.

As for the second and third stages, I'm curious really because I don't envision this being the *cleanest* pre in my collection - I have more modern/transparent pres for those duties. I quite like the idea of a bit of subtle distortion and saturation for amplifying instruments and giving things a bit of character (within reason of course).. I'm intrigued by the idea of introducing negative feedback, I'll read up on that.

Your advice about leaving the osc tube in place seems the best option - am I right in thinking therefore that all related components in that section starting with C16 (0.01uF) up to the osc coil and above the earth line can go? What about the trimmer cap, is this for hum control? Wiring is very messy so I think tearing down and rewiring so that the heater wires are properly twisted and more organised should help.

The tone circuit is the interesting point, I think I first need to identify which part of the circuit is responsible for adding that treble boost during recording but Im not sure which part is.

Hi Nick, well done on scoring your pair of M8s - they seem to be difficult to come by in the UK! There's an awful lot of info out there on modding the M7/M8 (as you've pointed out) although they often have a different set of valves and other such differences (the tone circuit in mine for example) and the 707 seems to be a little different, hence I'm not sure the established mods will be quite as straightforward. Hopefully given the amount of info there is on the m8, your life should be a bit easier!

 

[ruffrecords]

Thanks for your input Ian - so when we say relatively unchanged at the first stage, is that to say that all resistor and cap values can be left alone? The first 6AU6 has a 1uF electrolytic between screen and cathode whereas the modified diagram has a 0.1uF. How would changes here affect the first stage? How about the option of a DI for bass guitar - would this be inserted into the first stage or would it not work particularly well?

The 0.1uF would reduce the gain but it is not an optimum way to do it. You have two pentode stages each with a large gain. Far more than you need. Before you can decide where to put a DI input you need to decide how much gain you need and how to achieve it.

As for the second and third stages, I'm curious really because I don't envision this being the *cleanest* pre in my collection - I have more modern/transparent pres for those duties. I quite like the idea of a bit of subtle distortion and saturation for amplifying instruments and giving things a bit of character (within reason of course).. I'm intrigued by the idea of introducing negative feedback, I'll read up on that.

Triodes produce generally pleasing mostly second harmonic distortion. Pentodes tend to produce a lot more third harmonic distortion which sounds more edgey.

Your advice about leaving the osc tube in place seems the best option - am I right in thinking therefore that all related components in that section starting with C16 (0.01uF) up to the osc coil and above the earth line can go? What about the trimmer cap, is this for hum control? Wiring is very messy so I think tearing down and rewiring so that the heater wires are properly twisted and more organised should help.

I have marked up your schematic with red crosses on the oscillator components you need to remove. I would leave the heaters well alone. There is probably very little if any improvement to be gained by altering them.

The tone circuit is the interesting point, I think I first need to identify which part of the circuit is responsible for adding that treble boost during recording but Im not sure which part is.

I have also circled the frequency sensitive components that contribute to the EQ on recording and replay. You will notice the cathode bypass of the second 6AU6 is much smaller in record than in play. This is one source of treble boost. The other is the little 50p capacitor to the right of it that I have also circled. See the 1M and 2M resistors it connects to? These form negative feedback around the power stage. At high frequencies the 50pF starts to short this out and increase the gain from about 1.6KHz upwards.  All the circles components should be removed along with the 1M and 2M resistors. Replace the second 6AU6 cathode bypass cap with the 20uF used in replay.

Cheers

Ian

 

[untune]

Thanks again for your comprehensive reply Ian, this is a huge help - I've added your modified diagram to my notes. Given the large gain, is there any benefit in this instance of operating the either of the 6AU6 as a triode, or would that introduce more issues? (As an aside, out of curiosity - is there a reason that Akai/Roberts would have chosen such a high gain design or was it just common for the time?)

In regards to a DI I'd be looking to amplify a passive instrument level signal to a line level on the output. Passive bass output levels could vary but I remember reading something like 0.7v-1v peaks for normal playing. I'm not entirely sure what the output level/impedance is for this machine - the output jack interrupts the signal to the 8 ohm speaker but beyond that I couldn't say.  There's meant to be an 8ohm resistor in parallel to the VU meter in mine according to the schematic, but I can't find it in the circuit. I'm assuming that the switching between play and record ensures that the speaker is only in the circuit on PB but the switching scheme is taking a while to get my head around.

Thanks for explaining the EQ components - would removal of any of these components have any effect on the tone control, once the record/playback divide has been eliminated? Or should I be thinking about putting a Baxandall style control in that can be switched in/out? As mentioned I quite liked that little low end bump that the bass switch introduced during the 'record' phase.

I'll try to adjust the schematic and work on eliminating the rec/PB variations and simplify it into one circuit

Cheers
Lee

 

[ruffrecords]

Thanks again for your comprehensive reply Ian, this is a huge help - I've added your modified diagram to my notes. Given the large gain, is there any benefit in this instance of operating the either of the 6AU6 as a triode, or would that introduce more issues? (As an aside, out of curiosity - is there a reason that Akai/Roberts would have chosen such a high gain design or was it just common for the time?)

The large gain was necessary because they have a very small input signal. This is probably a quarter track machine running at 7.5ips or 3.75 ips. Doubling the speed doubles the head output and doubling the track width does the same thing. So a pro machine with half track heads at 15ips has 12dB or maybe even 18dB more output at the head. Add the fact that  to get a decently low noise level you need a step up transformer for a tube mic pre which itself adds more gain, you can see that we need a lot less gain than a tape recorder.

The 6AU6 works very well as a triode. That is how I used it in my very first tube mic pre design.

In regards to a DI I'd be looking to amplify a passive instrument level signal to a line level on the output. Passive bass output levels could vary but I remember reading something like 0.7v-1v peaks for normal playing. I'm not entirely sure what the output level/impedance is for this machine - the output jack interrupts the signal to the 8 ohm speaker but beyond that I couldn't say.  There's meant to be an 8ohm resistor in parallel to the VU meter in mine according to the schematic, but I can't find it in the circuit. I'm assuming that the switching between play and record ensures that the speaker is only in the circuit on PB but the switching scheme is taking a while to get my head around.

This is a relatively large signal so unless you want a fuzz bos you probably need to feed it in to a later stage. At some point you will have to decide if you want to convert this to 600 ohm output and to do that you really need to change the transformer.

Thanks for explaining the EQ components - would removal of any of these components have any effect on the tone control, once the record/playback divide has been eliminated? Or should I be thinking about putting a Baxandall style control in that can be switched in/out? As mentioned I quite liked that little low end bump that the bass switch introduced during the 'record' phase.

Cheers
Lee

Removing the EQ componets will not affect the tone control. That is the variable resistor and series capacitor hanging of the anode of the first 6AU6.

Cheers

Ian

 

[untune]

Hi Ian,

Of course, this makes perfect sense - I've been looking at the circuit with a view to making a standalone preamp for so long, I had overlooked the fact that it still has to function as an amplifier for the tape signal :-[

Based on your experience, would it be better to run one or both stages as triodes in this case? Does it have the same effect on the tone (the aforementioned edginess) when running a pentode as a triode?

How about a DI into the second 6AU6? In regards to output, a line out is more important than a speaker output so the transformer will probably have to be swapped. A Carnhill will probably be a good option, although I like Sowters too - I'm guessing the cost will be significantly more. Something like the 9600:600 VTB2291 for this application? or a 15K:600R?

I see now, noted with thanks. Would this now render the tone switch (SW4) ineffective, given that the negative feedback components have been removed? It appears the switch would work with the 100pF cap that has been taken out - or would I also remove a connection or move a cap elsewhere?

I'll start drawing up a new schematic when I get a chance!

Cheers
Lee

 

[ruffrecords]

You first need to work out how much gain you need for mic and DI inputs. Are going to record gnat farts at 100 yards with a low sensitivity ribbon mic? I not then you don't need a 1:10 mic transformer plus two pentode stage plus and output stage with a gain of close on 100dB.

70dB is the most you will ever likely need for a mic input and most times you'll need 40dB or less. A single triode 6AU6 will  provide  20 to 30dB of gain so with a 1:10 transformer at the front your fist stage has between 40 and 50dB of gain. Tha is why I suggested combining the second stage with the power tube and some NFB with a modest gain of 10 or 20dB.

The DI input could enter after the input transformer ans therefore have 20dB less gain.

For the output stage you will need a gapped transformer so something like the 2281 rather than the 2291.

Do you really need tone controls on a mic pre?

Cheers

Ian

 

[untune]

Hi Ian,

I had a think about this last night and made a few quick notes...

• As much as I want to realise my dream of composing something entirely from distant gnat farts, I'm going to have to put that one on hold for another project I reckon A max of 70db mic gain is more than enough, 60db max is probably more realistic - but the 70db would be nice to have as I do have a few old ribbon mics - however, most have been modified with new transformers upping them from 50 to 300 or 600 ohms, and as mentioned in the original post I have a handful of external transformers designed to bring lo-z ribbon mics up to something more reasonable. So let's say 60db.
• DI goals would be to bypass the mic transformer anyway and go straight to the preamp stage via a hi-z [1M] input (Bo Hansen DI if I need the mic input option although I'd have to invest in a phantom power box) and end up with a lo-z balanced out at line level. The REDDI thread had some interesting info and that would be good to shoot for so shall we say 16db of gain there. I like the option to overdrive things a little bit.

I like your idea of combining the second stage with the power tube, I think that's the way to go - although I'll probably need a little help in figuring out how to implement that. I've been reading up what I can but there are only so many hours in the day, unfortunately

Noted, thanks - I had a look at Carnhill's design guide and the 2281 is also ungapped (600:600 I think), whereas the 2290 is the gapped version of the 9600:600 - that the better option?

I know it sounds a bit daft having tone control on a pre, but I'm intending to run line level synths/effects and such through the line input too, to give a bit of saturation/breakup and a little bit of tone shaping would be a nice addition, especially since I have a tone knob and a bass/flat switch on the front panel already. A little like having EQ on your channel strip but a bit more subtle. Looks like the tone in this design is just a simple treble cut but I could always put something more interesting in and stick a bypass switch in!

Cheers
Lee

PS I've started a new schematic but work commitments have slowed progress!

 

[ruffrecords]

OK, 60dB sounds like a reasonable target.

My bad about the Carnhill transformers. I mistakenly thought the 9 changed to an 8 when you moved from ungapped to gapped but instead the last digit changes from a 0 to a 1.

The output tube is what in the UK is known as an EL84,a well known and much loved tube. Most of these work best looking into a load of somewhere in the region of 3K to 5K. Because the transformer carries dc current it needs to be gapped so it does not saturate. So the 2290 is probably the right starting point. This can be wired either 9K6:600 or 2K4:600. Difficult to say which is the best ratio to use. 9K6 is a lower oad but the anode will need to swing twice as far - but in a single ended transformer output topology it can swing up to twice the HT voltage so this may not be an issue. I have not designed this type of output stage before so I am not the best person to advise you on it.

Adding a cutting EQ is relatively easy. Adding one that boosts will require extra gain. In a mic pre, probably the most useful EQ is a HPF for getting rid of studio rumble of bass spillover.

Cheers

ian

 

[PRR]

> EL84,.. work best looking into a load of somewhere in the region of 3K to 5K.

5K (depends on V and I) gives far more power than a mike-amp out needs. (+37dBm!)

EL84 can be worked as resistance-loaded triode and deliver roughly 1/10th the power, which is still ample. R-loaded means no DC in the iron, saves much bucks. A hasty crayon-sketch suggests 0.2W (+23dBm) with 2.7K DR resistor and 5K AC loading. 250V B+, -5V bias, 30mA current, 167r cathode resistor. EL84 plate resistance near this point is 1.5K, so transformer has some ~~1K damping. A 5K:600 OT will show ~120r Zout, before NFB. >10Vrms output at clipping.

 

[untune]

Hi Ian,

Apologies for not following up, had a very busy schedule lately.

No problem on the transformer.  The Carnhills are relatively cheap and I'm sure the Akai OT (5k to 8R I think) could be used in another project. It dawned on me yesterday that I can't just wire straight into the power transformer (117v pri) which people often do with these units in the States...  I'll have to keep the original voltage selector and step down in place too, unless I opt for an external step down... I do have one somewhere. Any advantages to that?

I'm still in the midst of updating the schematic, removing the necessary components etc. Hoping to get it done in the next few days, time permitting. I was looking at inserting the DI input between the volume pot and 2nd 6au6. Also pondering if it might be interesting to make it switchable so it can be inserted at either the first or second tube stage.

In terms of the tone control I might go for a variable HPF, and rework the switch to bring it in or out.

I'll have to do some serious reading up in regards to the output transformer as it's all very new to me. Thanks for the info PRR, I just have to work out how it all applies I'm also wondering if I might be able to insert an output level control, so that I can overdrive the input and attenuate the output.

 

[pvision]

Hi Nick, well done on scoring your pair of M8s - they seem to be difficult to come by in the UK! There's an awful lot of info out there on modding the M7/M8 (as you've pointed out) although they often have a different set of valves and other such differences (the tone circuit in mine for example) and the 707 seems to be a little different, hence I'm not sure the established mods will be quite as straightforward. Hopefully given the amount of info there is on the m8, your life should be a bit easier!

Didn't mean to stomp all over your thread. I had read they were very similar. I might spin mine off into another thread as as not to complicate

The front panels of the 707 and M8 are very similar but that signifies nothing!

Nick Froome

 

[untune]

Didn't mean to stomp all over your thread. I had read they were very similar. I might spin mine off into another thread as as not to complicate

The front panels of the 707 and M8 are very similar but that signifies nothing!

Nick Froome

Hi Nick

Nah not at all, there are certainly a lot of similarities - I've studied the M8 info and there also seems to be a few considerable differences in regards to ins/out, the valves, front panels, even the knobs and VU meters seem to be wildly varied between the same models. Akai must have been refining things as it went along... I'm hoping to get my hands on an M8 at some point, and if there's any crossover between the two projects it makes sense to share the info

Cheers
Lee