Bush DAC90A - Adding Bluetooth

[DonaldStott]

Originally Posted by ms660

It was the first thing that caught my eye earlier on too, any chance of the Bluetooth specs? As I know virtually zilch about these devices.

This is the only thing I can find :-

KRC-86B V4.0

States that it "... is based on the CSR 8630 bluetooth chip": -

CSR 8630

What I have found is that the audio quality of this Bluetooth board when connected directly to the top of the volume control of the DAC90A is perfectly acceptable. Some users have reported hearing a high pitched 'digital noise' coming from the power supply (?) but this is not something I've experienced.

I'm edging closer to implementing this project without any impedance matching transformers but just having 100R resistors in series with the L and R channels which are combined at the output stage?

[ms660]

Cheers for the links.

Lawrence.

[DonaldStott]

Originally Posted by ms660

Cheers for the links.

Lawrence.

This link has much more in the way of technical information - CSR8630

[ms660]

I Googled the Bluetooth module, not a lot of info about the output so far, might try again later on.

Lawrence.

[wd40addict]

Have a read of this thread:

https://www.vintage-radio.net/forum/...d.php?t=168931

Max output of the KRC-86 is 900mV rms with source volume at max, Zout 390R. I'm sure this is sufficient to drive the DAC90 amplifier. I don't see any need for impedance matching

[ms660]

Originally Posted by wd40addict

Max output of the KRC-86 is 900mV rms with source volume at max, Zout 390R.

Cheers for that.

Lawrence.

[wd40addict]

The problem with direct substitution of KRC into a valve signal path can be the fitted 100k resistors on the I/O. Remove them and the problem goes away.

Obviously any Bluetooth circuit fitted to an AC/DC type chassis must be safely tucked inside the cabinet, but then Bluetooth itself provides the safety isolation to the audio source.

[mark2collection]

Having used a KRC module in a modern solid-state amp, and a Bush VHF61 for someone in the family :-

1. Ensure the KRC PSU ground goes to the sets PSU ground, and not a convenient place on the chassis.
2. Only ground one end of any signal leads, keep the other end insulated/isolated from other ground connections.
3. Ensure the KRC is mounted so the on-board antenna is pointing away from analogue signal leads/controls.
4. If you use the 'connected' confirmation LED beacon supplied with the KRC module, and extend the wiring, twist the wires together and keep them as close to the chassis as possible.

This from my experience kept digital noise & hum to an absolute minimum. In the case of the VHF61, it was completely noise free & sounds very good. The solid state amp took a bit more work, but now sounds excellent, also noise free.

I had flirted with the idea of 'modding' the DAC90a sat in my workshop, using the KRC 'loop-through' feature for standard analogue broadcasts, but when called for, the KRC module would cut the analogue broadcast off & send audio to the sets volume control. Best of both worlds, the set still works to factory spec, but has the added feature of Bluetooth.

PSU wise, the picture in my head would be via the sets PSU, safely dropped, smoothed further, regulated, protected etc.

Cannot remember the values used in the VHF61, the KRC stereo output had to be mono'd via a couple of caps & a set of resistors, else if you only had one output connected on the KRC & listen to the Beatles, you'd only hear either vocals, or the instruments, depending which channel was connected(!)

The set of resistors was to ensure the output of the KRC at full volume via an iPhone matched the levels of a strong signal on standard broadcast, this saved being deafened when changing from say, FM, to 'Gram-o-tooth'

Mark

[DonaldStott]

Originally Posted by wd40addict

Have a read of this thread:

https://www.vintage-radio.net/forum/...d.php?t=168931

Max output of the KRC-86 is 900mV rms with source volume at max, Zout 390R. I'm sure this is sufficient to drive the DAC90 amplifier. I don't see any need for impedance matching

Thanks for this and the link back to the previous Thread - I remember reading this but it didn't make much sense at the time, now it does (??) but it's at the limits of my technical understanding! I'll try again to digest the contents and incorporate the recommended changes into my own design? As mentioned earlier my current configuration, which does work to my satisfaction, just has 100R resistors in series with the L and R channels which are combined and output to the top pf the volume control? I'm not sure to what extent this arrangement is valve friendly or not?

I measured the output current of the unmodified KRC-86 with my DMM and only got 30-32mA, peaking at 35mA? Of course this is before the removal of the 100k on-board resistors.

I am, however, starting to spot a recurring theme i.e. there is no need for impedance matching transformers?

[wd40addict]

There's no need for the 100R resistors either, the internal resistance of the analogue switch does that for free. The only decision is whether to feed the audio from the detector diode through the switch or to use the low impedance of the module when in Bluetooth mode to overdrive the the existing signal.

Your use of overdrive sounds like it's not enough so try removing the 100k I/O resistors and switching the radio audio instead.

Removal of the 100k resistors will make no difference to current.

[DonaldStott]

Thanks WD - I've been looking at this circuit diagram from a Colin Boggis Post back in March 2020:-

https://www.vintage-radio.net/forum/...6&d=1583078374

I have everything to the right of the KRC-86B designed, tested and working with AC coming from the DAC90A on/off switch-volume control but not sure how to amend the Inputs and Outputs in line with your recommendations? The KRC-86B also has two AGND (audio ground) connections not shown in the circuit diagram?

You also mentioned this:-

Originally Posted by wd40addict

... the internal resistance of the analogue switch does that for free. The only decision is whether to feed the audio from the detector diode through the switch or to use the low impedance of the module when in Bluetooth mode to overdrive the the existing signal.

A bit unsure as to what you refer to as 'the switch'?

[Colin Boggis]

This is a repeat of a previous post I made, it may help you:
I recently restored a very tidy DAC90A and added bluetooth. As I didn't want to modify the wavechange switch or drill holes for an extra switch, plus I wanted it to be possible to easily convert it back to original design I hit upon an alternative way to do it. I made up a sub- assembly with a 5V power supply (transformer, rectifier, regulator) and a long lever micro switch. I mounted this on the dial frame using existing bolts, such that the lever could be bent to be activated when the tuning pointer was at the far left hand end. I made another small pcb to accomodate a mini changeover relay and a Bluetooth module. The only other wiring change made was to disconnect the detector to volume control lead and reconnect it to my new pcb. Another lead went from the pcb to the volume control. When the relay was not activated the detector was connected through to the v/c, and when activated the relay changed over to feed the BT module through. I also placed a 1M resistor from the "bottom" of IFT2 secondary to pin 3 on V2 to provide a load to the detector whilst the relay was activated. This BT solution was reversible.

[Colin Boggis]

I just realised you already read my article! The switch through which the audio was fed in my arrangement was the miniature relay contacts.

[Gabe001]

Donald, what WD refers to is automatic muting of the radio when the BT is paired. The BT module you're using has the ability to do that (switch the radio and BT inputs) but the 100k resitors need removing as shown in my cossor thread (see picture).

The alternative is to go onto a silent part of the AM band and have the BT output override the radio signal and in that case the BT module needs no modifications.

The biggest problem I find is finding a good mounting point. I don't want things knocking about and I'm not convinced of the merits of hot glue in sets that get hot. Where, and how, are you going to put your BT board and power supply in the DAC90a?

[wd40addict]

The analogue switch is the CD4052 mounted on the KRC PCB. It selects automatically between the output from the CSR BT device and an incoming audio signal.

The audio grounds just connect to the main grounds on the PCB.

Initially I'd just connect it up as per the attached diagram and see what happens, but you will have to remove the 100k resistors otherwise the poor detector will only see a 25k load.

It could still be that the performance is worse than overdriving, but you won't know till you try it.

Note the resistors are tiny surface mount devices, if you are not confident of removing them stick with your current overdriving arrangement!

[wd40addict]

Originally Posted by Gabe001

The alternative is to go onto a silent part of the AM band and have the BT output override the radio signal and in that case the BT module needs no modifications.
?

Not entirely true, when in radio mode, with the resistors in place, the detector will see a 25k load in normal radio use as opposed to the 500k it's expecting. Bit low for a valve circuit, but you might get away with it.

[DonaldStott]

Thanks for all the recent helpful Posts - seems clear that I have a number of options to try and that there are several different ways to configure the KRC-86B Bluetooth board.

Just getting to grips with concepts such as impedance matching, loop-though and overdriving, amongst others, but I like to understand what I'm doing rather than just following a circuit or layout.

I'll try and report back as best I can on how I got on with the various options.

(In the interim, although I'm not phased by removing the surface mount 100k resistors I'd welcome any practical advice on how best to do this? Don't want to end up with a batch of botched BT boards!)

[wd40addict]

There are different ways of doing this, the best is obviously thermal tweezers. Alternatives are two small irons used simultaneously, or where a component is as small as this one hopping from end to end with a well tinned bit until it starts to move and can be flicked off with the iron.

Another thought on your overdriving: if you're doing it with the 2*18k shown in the drawing you posted you will get some radio in the background as you're not getting the full benefit of the module's low output impedance. Try bypassing them. If you have fitted the 47uf cap shown then the positive terminal should connect to the radio and negative to the KRC, not as drawn. This is because the KRC's output is centred on 0v, whereas the DAC90A volume control bottom end is biased at +1v.

[DonaldStott]

Thanks WD - can't see me splashing out for some thermal tweezers any time soon as I don't do much in the way of surface mount work! I'll try using the smallest solder bit I have and proceeding as you suggest - wish me luck!


The link I Posted was to the circuit diagram by Colin Boggis and not my own - I don't use the 2 x 18k but 2 x 100R resistors in series with the L and R channels which are combined and output to the top pf the volume control? Neither do I have the 47uF cap and not sure why this is required but note your advice about polarity?

[wd40addict]

The cap will stop a bit of scratchyness as the volume is adjusted in BT mode and will stop any effect on the cathode bias of V2 & V3 when the volume is set low in BT mode. 47uF is rather large though