Yes, at the minimum, we'd want a high efficiency schottky diode (0.28v or smaller forward voltage drop) between the two V- pins.
yes, it's in there (for both V+ and V-). D2 for each rail in the BOM
note also that the tube buffer version does not significantly compromise PSRR and tube sonics because the output is taken from the anode rather than the mu point of the CCS. the mu point would offer lower output impedance but that would compromise PSRR and tube sonics; the lower impedance is not really needed given the 6922's very low Rp and the CCS feed.
Last edited:
Indeed. It occurred to me that a schottky diode with reservoir and high frequency bypass capacitors dedicated to the input stage V+ would be an alternative to the datasheet circuit, but probably not as good as a dedicated V+ supply. It would be much simpler, and more foolproof, than a dedicated V+ supply.
Does the 3886 suffer from the same flaw? I haven't seen any 3886 circuits employing separate V+ for the input circuit, yet the option is available. Can it be isolated with a diode or resistor? I'm going to give a 3886 amp a shot I think - all I need are the chips and a few electrolytic capacitors.
I haven't noticed the V+ pin for the input stage being brought out separately on the LM3886. Which pin is it - the internal schematic only shows a single V+ net?
I'm one of those who blew up my first TDA7294 due to adding resistors between the two supply pin pairs. However I reckon the negative one is the biggie since the substrate is connected to the signal V- (not the power V-). I'm about to do a build in bridged and I'll try an inductor on the positive supply only.
I'm one of those who blew up my first TDA7294 due to adding resistors between the two supply pin pairs. However I reckon the negative one is the biggie since the substrate is connected to the signal V- (not the power V-). I'm about to do a build in bridged and I'll try an inductor on the positive supply only.
The parasitic diode-like effect of the substrate is on V- and connects to tab; so, that is the breakable part, but fortunately it is not as efficient as an MBR735 (for example). As for the V+ it is more tolerant, but I do not know how much.
In any case, the amplifier board power decoupling caps are small signal, so if you're going to use a series element, then "at/upon the amplifier board" if "in series to the dc umbilical" is a really premium spot. It works fine with ordinary split rail (no need for quad rail). The very easiest to apply for refit purposes is an MBR1645; however, many other options for series element do exist, albeit with a somewhat different amount of labor involved.
In any case, the amplifier board power decoupling caps are small signal, so if you're going to use a series element, then "at/upon the amplifier board" if "in series to the dc umbilical" is a really premium spot. It works fine with ordinary split rail (no need for quad rail). The very easiest to apply for refit purposes is an MBR1645; however, many other options for series element do exist, albeit with a somewhat different amount of labor involved.
the current version of this thread's P2P build has HF and LF bypasses for each power pin, a separate feed for each power pin and schottky diodes (the two D2 devices on the schematic) between Vp and Vs for each rail. it's agnostic whether there are separate Vp and Vs supplies on the PSU end of the power cable. if there is a concern about cable failure then the D2s could be moved to the relevant chip pins.
if the schematic is unclear as to any of this please let me know so i can improve it.
all that said, we wont know how much good the isoltaion will do until we try it as the datasheet contains very little PSRR data.
if the schematic is unclear as to any of this please let me know so i can improve it.
all that said, we wont know how much good the isoltaion will do until we try it as the datasheet contains very little PSRR data.
Last edited:
unless the LM3886 connection diagram (fig 2) is wrong in the datasheet there is just one V+ and one V- (both pin 1 and pin 5 are identified as V+) and in any event it's the V- PSRR that is relatively poor. V+ PSRR is decent enough for most purposes (maybe not good enough for SMPS).
the dual version (LM4780) has a multitude of power pins but no documentation on whether there are differences among them rather than simple parallel
Last edited:
on a separate topic -- does anyone know what it would take to drive pin 11 of the TDA7293 from an external source? my main question is the impedance/current aspect. with a good parallel 6922 driver fed into a bridged unity gain (well, OK, 2x, actually) chipamp i would have no problem swinging each chip rail to rail in the 20 to 24 volt range.
with 6 TDA7293 in a BPA topology that could deliver a gigantic amount of clean power into 4 ohms (each chip seeing a 6 ohm load).
with 6 TDA7293 in a BPA topology that could deliver a gigantic amount of clean power into 4 ohms (each chip seeing a 6 ohm load).
Last edited:
I've never tried feeding a 7293 slave output stage from something other than a master 7293 preamp output, nor monitored/recorded the voltage and current.
So I also do not know if full protection is still warranted by the output stage proper (probably is, though).
BTW, you can circumvent PSRR/CMRR issues in non-inverting mode by running the frontend from a higher (and well filtered) supply modulated with the input signal (derived via a low valued resistive divider from output). This is a big advantage of the seperate power pins.
So I also do not know if full protection is still warranted by the output stage proper (probably is, though).
BTW, you can circumvent PSRR/CMRR issues in non-inverting mode by running the frontend from a higher (and well filtered) supply modulated with the input signal (derived via a low valued resistive divider from output). This is a big advantage of the seperate power pins.
"...
The [LM4780] Design Support Group responsed on 14 September 2004 'Here is the connection of each die:
Pin Connection
1 N/C
2 V- (Vee) Down bond to the DAP of the package
3 N/C
4 V- (Vee) to die A 1st V- pad
5 V+ (Vcc) to die A 1st V+pad
6 N/C
7 Output from die A pad
8 V+ (Vcc) to die A 2nd V+ pad
9 V+ (Vcc) to die A 2nd V+ pad
10 V- (Vee) to die A 2nd V- pad
11 V- (Vee) to die A 2nd V- pad
12 GND to die A GND pad
13 N/C
14 Mute to die A Mute pad
15 INA- to die A INA- pad
16 INA+ to die A INA+ pad
17 V+ (Vcc) to die B 2nd V+ pad
18 V- (Vee) to die B 2nd V- pad
19 GND to die B GND pad
20 Mute to die B Mute pad
21 INB- to die B INB- pad
22 INB+ to die B INB+ pad
23 N/C
24 N/C
25 Output from die B pad
26 N/C
27 V+ (Vcc) to die B 1st V+pad
Internal Downbond to DAP from die B 1st V- pad
..."
by swinging down i gather this implies sinking (rather than sourcing). it makes a difference because some tube/valve topologies (such as cathode follower) source current ok but don't sink current well. but 2 to 3 ma is not a big deal for parallel 6922 (2 or 4) active loaded in common cathode. probably worth a try. at worst i blow the chip.
I read this somewhere but couldn't find a source. I saw a block diagram that I can't find now. Anyway, pin 5 is connected to V+. Does anybody know what pin 5 is for?
I have a chassis, heat sink, dual mono power supply (+/- 30 volts @ 6 A each) ready to go and need some amplifier boards. I was going to whip up something simple. I would like to know what pin 5 is for, though; even though I'm just going to hook it up.
Quad Rails with the TDA7293/4/5/6 serves to decrease reliance on the bootstrap cap, and if the voltage is different enough to reduce the max power to ~40W, then you wouldn't need the bootstrap cap.
Well, TDA7294 at 40W versus the LM1875's 22W is approximately 3db difference in speaker output, so I never explored the quad rails thing with TDA7294 to get sound not quite as good as the LM1875, just 3db louder.
Well, TDA7294 at 40W versus the LM1875's 22W is approximately 3db difference in speaker output, so I never explored the quad rails thing with TDA7294 to get sound not quite as good as the LM1875, just 3db louder.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.