Agreed, the valve needs DC current to flow in order to operate. I assume he ment to say that when using an IT or OPT the DC element cannot pass from the primary to the secondairy coil without directly connecting them with a wire. In this case you can say the IT / OPT can be used to block DC voltages.
In my opinion there are two ways to couple between amplifier stages: AC and DC. With AC coupling you can use a capacitor or a transformer. With DC coupling this would be a wire, or at most a resistor.
On a side note ... It is possible to replace the capacitor with no capacitor. I have been working on this issue for the last couple of years because I wanted to get rid of coupling caps, and I can confirm that it is actually possible. As an example, my current test amplifier is a AB1 EL84 amplifier with a ECC88 LTP front-end. The amplifier is DC coupled from input to the output transformers. All operating voltages are under normal conditions: Input is at 0V, EL84 grids are around -12V, B+ is 300V, the output power is 8.6 watts and there are zero (0) coupling caps in the circuit.
Peufeu, I do agree COG are much better, but not available in large values and small cases . A lot of newer LDOs are designed for local digital supplies and to work with the larger value MLCC caps 1-10uf in the smaller cases so the supplies can be near the device. Which LDO were you using, I was on about the Vout pin, for the 3 pin devices. Also what cap make if you know I will have a look at the specs.
Cheers for the scope shots, all we need to do now is determine how much force a Twangggg has
Unfortunately the firm where I could play with these things (Cobham Blackburn) and where we had a vibration table (a real beast) has been downsized (200+ to 20!) and due to a virus on my old PC a lot of my stuff has gone.
The tests were done on all SMD boards with 0402 as the smallest components, the supplies were large linear for system main voltages, SMPS for the on board 5V & 3V3s these were then fed to local LDOs for all the silly voltages, 1.8V, 1.2V 1.0V, mainly digital with some codecs.
Depending on the product though most PCBs in a case in a domestic environment have a pretty easy time, constant temperature and relatively vibration free, so these issues shouldn't normally be a concern, though on oscillators we did get better results with COG, so for really critical situations the next to the pin decouplers should be low value COGs.
Cheers for the scope shots, all we need to do now is determine how much force a Twangggg has
Unfortunately the firm where I could play with these things (Cobham Blackburn) and where we had a vibration table (a real beast) has been downsized (200+ to 20!) and due to a virus on my old PC a lot of my stuff has gone.
The tests were done on all SMD boards with 0402 as the smallest components, the supplies were large linear for system main voltages, SMPS for the on board 5V & 3V3s these were then fed to local LDOs for all the silly voltages, 1.8V, 1.2V 1.0V, mainly digital with some codecs.
Depending on the product though most PCBs in a case in a domestic environment have a pretty easy time, constant temperature and relatively vibration free, so these issues shouldn't normally be a concern, though on oscillators we did get better results with COG, so for really critical situations the next to the pin decouplers should be low value COGs.
> Which LDO were you using
LP2985 3v3 version.
> Also what cap make if you know
Hum... output cap is some 1206 10µF X7R... I don't remember the manufacturer. Not important anyway as the output capacitor piezo effects were completely buried in the noise floor (the LDO controls the output voltage).
It's the reference filter cap which matters, the manufacturer spends half a page of datasheet talking about it, for a reason :
> all we need to do now is determine how much force a Twangggg has
Yeah that's the problem of "when I do an unknown amount of stuff, other stuff happens"...
LP2985 3v3 version.
> Also what cap make if you know
Hum... output cap is some 1206 10µF X7R... I don't remember the manufacturer. Not important anyway as the output capacitor piezo effects were completely buried in the noise floor (the LDO controls the output voltage).
It's the reference filter cap which matters, the manufacturer spends half a page of datasheet talking about it, for a reason :
> all we need to do now is determine how much force a Twangggg has
Yeah that's the problem of "when I do an unknown amount of stuff, other stuff happens"...
Cheers, good data sheet.
Had a look around some circuits (from a variety of sources) and data sheets form various manufacturers and the majority have bog standard X7Rs in this position...size and cost come into play here for commercial designs I'm afraid. But thanks to your research I will recommend that a COG or plastic film SMD cap would be much better in this position, and any circuits I do I will use such a cap in future.
Again thanks for the information and scope shots, very educational.
Had a look around some circuits (from a variety of sources) and data sheets form various manufacturers and the majority have bog standard X7Rs in this position...size and cost come into play here for commercial designs I'm afraid. But thanks to your research I will recommend that a COG or plastic film SMD cap would be much better in this position, and any circuits I do I will use such a cap in future.
Again thanks for the information and scope shots, very educational.
If you do not parallel caps, there are no antiresonance peaks.
C0G ESR here : SimSurfing
Actually this high-Z REF node is a weak spot (it is high-Z because the regulator uses as little bias current as possible for its reference), and this is also true for high-Z nodes in many other audio circuits... all kinds of crap can capacitively couple into it.
New regs are out like ADP151 which do not need a filter cap, much better solution IMHO.
Another example, I used a water-soluble flux which is really excellent, but leaves vaguely conductive residue... after cleaning, enough of it stayed trapped under the SMD capacitor to cause leakage problems...
Anyway, back to those big audio film caps, they are large, so they are a prime target for picking up noise by capacitive coupling with a nearby piece of metal that has an AC voltage on it, like a PCB track, or another capacitor nearby. The impedance of the node they are connected to matters ! It can be a good idea to put a grounded shield around them.
C0G ESR here : SimSurfing
Actually this high-Z REF node is a weak spot (it is high-Z because the regulator uses as little bias current as possible for its reference), and this is also true for high-Z nodes in many other audio circuits... all kinds of crap can capacitively couple into it.
New regs are out like ADP151 which do not need a filter cap, much better solution IMHO.
Another example, I used a water-soluble flux which is really excellent, but leaves vaguely conductive residue... after cleaning, enough of it stayed trapped under the SMD capacitor to cause leakage problems...
Anyway, back to those big audio film caps, they are large, so they are a prime target for picking up noise by capacitive coupling with a nearby piece of metal that has an AC voltage on it, like a PCB track, or another capacitor nearby. The impedance of the node they are connected to matters ! It can be a good idea to put a grounded shield around them.
No, what I meant was a copper nickle alloy. I work on cars and have recently found a new type of brake line. Its a copper nickle alloy that is white like aluminum. Very flexible and doesn't corrode, even in a car when exposed to weather. It would be much less expensive than silver and its anti corrosion properties would not have been lost on the Russians in military applications. I'll test is conductivity tomorrow and let you know. Maybe Anatoliy would know?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.