thanks for the updated schematic, this one indeed makes sense, can you point out which component value should I change in case of 8 x TDA? thanks!
thank you very much, this is the 8 chip adapter I have so would like to give it a try using your schematic (it is an adapter for TDA1541A direct replacement)
An externally hosted image should be here but it was not working when we last tested it.
Here's the variant of the filter/IV stage suited to 8 DACs.
In preparing this one I discovered there's a component error on the 4 DAC version, which I'll correct shortly.
I've also realized I didn't specify the type of LED - its a green one with 2.1V forward drop. The idea is that the DAC output voltage shouldn't rise above 3.5V (which is the upper limit of its compliance range). If it does, the resistor R15 should be increased to lower the voltage at the BC807's base.
In preparing this one I discovered there's a component error on the 4 DAC version, which I'll correct shortly.
I've also realized I didn't specify the type of LED - its a green one with 2.1V forward drop. The idea is that the DAC output voltage shouldn't rise above 3.5V (which is the upper limit of its compliance range). If it does, the resistor R15 should be increased to lower the voltage at the BC807's base.
Last edited:
thanks, I'll give this a try as soon as I gather the components, I'm really curious about this DAC; what does BREF mean and is it necessary for the LED and C29 to be present only on Left channel? I see these parts missing on R channel; I see R14 floating on the right channel, do I have to connect this BREF to something?
Bare in mind this chip has been out of production for many years... 10!
The chance that old stock of such quantity is available is slim - far more likely our Chinese friends have created a simple ladder dac and printed the logo on top. They are doing this for sure with the TDA1387. I threw 10 out a few days ago, 5 had the same logo as in your picture and 5 had the Thailand text broke up around the logo.
People also said they were being reclaimed from old AWE cards. Firstly, that's a lot of AWE cards, secondly AWE cards go for a lot of money now. Doubt they are being butchered to supply a <$1 dollar chip!
Hey 'Braxy!
Basically a ~4dB rise from ~11kHz to 22kHz
Most use a 3db peak at 20kHz with the q selected to match the rising response of the sin c curve.
I recall CarlosFM made a simple, elegant 3rd order filter (NOS needs heavy HF filtering) with a 3dB peak at 20kHz around a fast video opamp ... neat!
Cheers,
Jeff
Last edited:
BREF is the reference voltage supplied to the base of the common-base transistor (BC807). Its generated by the LED+R15 being fed current via R16. Only one instance of the reference generator is needed hence R14 connects to C29 -ve. Put another way, connect both instances of 'BREF' together.
Are you aware how much engineering goes into the creation of a DAC chip? I'm sure even Chinese IC designers have better things to do than 'recreate' an ancient 1u chip which then sells (on Taobao) at 35 to the USD.
What voltage are you targeting for BREF?
If I understand correctly (a dicey proposition!), the BREF value is such that we get 3.5V on the bc807 emitter (pin2). To achieve this, we need to set BREF accordingly. I'm not sure I understand how to determine that value.
Taking a stab at it based on (likely poor understanding of) your comments from our email conversation about the J327: For the bc807: am I right to look at the Base - Emitter On Voltage (V_BE(on)) on the datasheet to determine this value? Which looks like like the max is -1.2 Volts. So -1.2 = BREF - 3.5 -> BREF = 2.3V?
But, as drawn on the schematic above, assuming my math is correct, looks like BREF ought to be around 2.84V.
I'm sure there are several errors in my thinking above.
Doesn't it bother you that X7R capacitors (especially high-valued small-sized SMD versions) are quite nonlinear (as well as lossy and inaccurate)?
That's almost correct - the targeted value is to be 1.5V below the positive supply voltage on the output of the DAC chips. This is because the DS for the TDA1387 specifies the output compliance range as running from 0V to 3.5V when the supply is 5V.
In between the DAC output pin (pin6 & 8) we have the BC807 base-emitter junction and a small valued (under 2ohm) emitter degeneration resistor. (The DCR of the inductors is well under 1ohm so I'm neglecting them.) The former can be assumed to be 0.65V and the resistor voltage drop will be at most 5% of this value. So let's go for 0.69V as the difference between BREF and the DAC's output. It means we need 1.5V+0.69V = 2.19V below the positive supply for BREF. At 5V this means 2.81V.
That V_BE figure is very much worst case (500mA Ic) and not representative of the usage here where we have around 20mA for Ic.
Sounds close enough, looks like your math was pretty good
The issues with classII ceramics I'm not unaware of. Yes they're nonlinear with voltage - but the voltage variation they experience here is of the order of tens of mV max. I've studied the variation of bias plots over on TDK's website and for 1206 size its definitely not an issue. Small than that, quite possibly but I'm not yet recommending smaller than 1206 in this application.
My main concern with X7Rs is their fairly strong temperature coefficient. Inaccuracy is addressed by choosing better tolerances (10% is available) and selecting with my LCR meter. DIYers are free to choose closer tolerance caps.
As regards losses, ESRs don't look to be any worse than film capacitors? There is microphony too but its not been a problem in practice, having built many variations of this kind of filter.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.