LT1166 alternative
Hi Mason,
The AP0016 from Pioneer is a good alternative and housed in a 14 pins DIP.
Look here for more details.
edit: What about a group buy?
Cheers,
E.
Hi Mason,
The AP0016 from Pioneer is a good alternative and housed in a 14 pins DIP.
Look here for more details.
edit: What about a group buy?
Cheers,
E.
Last edited:
It looks like the LT1166C is available in lead-free SOIC under the number LT1166CS8#PBF for $3.25 in small quantities. I went to the LTC site and tried to order it and it seemed to let me. There are 3 versions listed on the LT1166 page, the PDIP, what appears to be the older leaded SOIC, and the #PBF version. It would not let me proceed with the first two. I also noticed on one of the pages I was led to by a search of the LT1166 that the first two of those three had the long lead time and high minimum order. I may have also seen a typo that suggested the orderable one is an LDO. So be careful.
In the business, many vendors have dropped the leaded versions of their product. I see this all the time in my day job. I cannot tell you how many times I have cursed ROHS. Looks like in this case LTC dropped the DIP completely, not replacing it with a ROHS version.
I know SOIC is a PIA for a lot of us, but at least the part may still be available. In the past, I've called the LTC people and they have been very kind and helpful. Try giving them a call.
Cheers,
Bob
Well, it has 50 mils pitch between the pins and no pad underneeth.
With a modern soldering iron and a good tip is it not that difficult. I am thankful it is not a BGA.
With some practice, can you actually desolder these with an ordinary heat gun. You shall use a nozzle with ~6mm diameter, but you need to make an extra hole in the side of the nozzle to get down the airflow over the board. Use some tweezers to lift the component.
I am, but at $20 a pop (which, most likely, turns into $40 during flying and arriving on this side of the pond) it's a nono for anything but a repair job.
If anybody can source genuine PA0016s at a reasonable price and is willing to organize a group buy, I am all in.
... and if you have soldering tweezers with the right attachment, you can lift them in one step. I have the Weller WMRT tool with the RTW4 attachment, and 8-pin SOIC is a breeze. I've even removed 14-pin SOIC op-amps with a little patience and a gentle hand.
PA0016 vs LT1166
Hi Bob,
You will be amazed that a far simpler chip (16 trannies opposed to 43 trannies) does a far better job.
Cheers,
E.
Hi Bob,
You will be amazed that a far simpler chip (16 trannies opposed to 43 trannies) does a far better job.
Cheers,
E.
My reasoning works a bit differently.
A good RF filter attenuates the RF.
A good RF filter does not attenuate the AF (audio frequencies).
If you can achieve both aims then you are doing really well.
On that basis I set the RF filter as low as possible without any audible impact on the AF that I want to pass through.
I generally adopt a passive RF attenuator of 680ns (1k & 680pF or 100r & 6n8F or whatever), such that I cannot hear the difference.
I see recommendations that the RF be set between 100ns and 1.5us to suit the listener and the equipment. I would go much narrower than that, try from 330ns to 1us. This is a very cheap experiment to carry out.
Last edited:
Hi Pooge,
The fairly high cutoff frequency of R11/C3 reflects the fact that it is there to suppress the possibility of HF oscillations brought on by rail feedback from the output transistor to the collector of the pre-driver and driver transistors. As noted in the section, the combination not only attenuates HF feedback, but also acts as a Zobel notwork on the rail for the output transistor, decreasing the Q of any resonance thereon. Most of the effects of concern are at frequencies of 10MHz and above, so the 1.6 MHz corner for the network is fine. Having said that, there is no reason why a larger value of shunt capacitance cannot be used for C3. In fact, I usually parallel it with an electrolytic. I will also sometimes put such an RC network in the rail path to the predriver as well. Since the current load is smaller, I might choose to use a 10-ohm resistor in series there.
Cheers,
Bob
Hi Bob,
Set aside I'm not convinced of any real world benefits of this (I was never able to determine any measurable effect of such RC snubbers, at least for audio purposes, high speed digital is a different story), the sad news is that unless wirewound (and unfortunately also inductive) or carbon composition 1 ohm resistors are used, they will blow after a very few power on cycles. Been there, done that, reasons are the peak current and peak power dissipation in the resistors.
Discharging some 10J of energy from the reservoir caps into an electrolytic C3 of say 100uF, at power on, with a time constant of a few 100s of microseconds, will lead to peak currents in the 10s of amps and peak powers in the 100s watts. Practically, for about one time constant (say 100uS) the power supply voltage drops on the snubber resistor only, the C3 electrolytic acts as a short. No film resistor will survive such, other than by pure chance.
Only a 0.1u ceramic with 1ohm film is fine, time constant are to short to blow the resistor. But then again, I'm questioning any measurable benefit of such a cell, for audio purposes.
Hi Waly,
First of all, we are talking about RF, not audio. Parasitic oscillations, especially in Triples, are an RF issue. But if you have them, it will quickly become an audio issue one way or the other.
If you can get a 60Hz power amplifier power supply to come up in 100us, you are a better man than I am. Think about how long a quarter cycle of 60Hz is. Measure the slew rate at the main rails of your power amplifier and see what it is.
It is also remarkable how much peak current many types of resistors can withstand, but it is certainly no big deal to put in a small wirewound resistor (not a big sand-cast one) - its inductance will help a bit anyway. A 3W metal oxide resistor will also do just fine.
Cheers,
Bob
Hi Bob,
You lost me with the RF explanation. Of course parasitic oscillations in triples are RF (as most of the parasitic oscillations) however I don't think this RC snubber has anything to do with (or helps alleviating) such instabilities. If adding such snubbers stops parasitic oscillations, then I think it can be safely assumed there's something deeply wrong with the designs' global or local stability (and/or PSRR, etc...). I don't think the power rails are the best place to cure or correct any instabilities or design flaws.
You are assuming the reservoir caps charging from the rectifier, as in the simplest power supplies. In fact, many power amps have switching mode power supplies (including PFC pre-regulators, a notable example are the Halcros). Such implementations are not switching on the power by flipping the mains switch, but by digital controlling the switchers, hence a potential large dv/dt.
Anyway, while carbon volume resistors are the standard for surge resistant resistors, 3W metal oxide resistors may do, never tried such big fellas. I though still find hard to justify the cost and the board space for such snubbers.
Thanks Bob. I'd seen ripple filters there. Just wanted to make sure the numbers were correct.
Wally,
The SMPS I've used have built in slow starts for turn-on. The rate can vary but I haven's seen one exceed 10V/ms.
Hope this helps
-Antonio
Hi Waly,
I have to disagree about strategic use of things like snubbers. If a snubber is placed in an amplifier it certainly does not suggest that there is something wrong with the design. It is fair to say, however, that a snubber can sometimes reduce the likelihood of HF parasitic oscillations due to a less than optimum layout. Unfortunately, there are unavoidable inductances in the real world that come into play, and which are often not modeled by SPICE simulations. You would be very surprized how many unintentional oscillator topologies can be formed by even small inductances, and these are more likely to become active with fast transistors. It is also often the case that these oscillations only occur in bursts under certain combinations of voltage and current.
The proper use of snubbers to kill the Q of oscillators can actually improve amplifier performance. Check out my JAES article on my MOSFET power amplifier with error correction. There I decribe many of these things and show how the use of gate Zobel networks allow the use of smaller gate stoppers, which tend to kill the inherently high ft of MOSFETs. Its available on my website at CordellAudio.com - Home.
I also suspect that there are some amplifiers out there that are occasionally breaking into parasitic oscillation bursts that affect the sound, sometimes in a subtle way, and in ways that may not show up in conventional bench tests.
Cheers,
Bob
blown resistors
Hi Bob,
There is some truth in Walter's comment. The supply rails of the PGP front-end, for example, were filtered by means of a 100uF cap and a 10 Ohms metal film resistor, rated at 1/4W. (see: Front End PCB Schematic R1 & C2, in the upper-right corner and R72 & C29 in the lower-right corner)
After powering up a few times, these resistors were blown. Admittedly, they were 10 Ohms instead of 1 Ohms. Nevertheless, I think we should not underestimate the peak power during startup, even in case of very small resistors, like 1 Ohm.
Cheers,
E.
Hi Bob,
There is some truth in Walter's comment. The supply rails of the PGP front-end, for example, were filtered by means of a 100uF cap and a 10 Ohms metal film resistor, rated at 1/4W. (see: Front End PCB Schematic R1 & C2, in the upper-right corner and R72 & C29 in the lower-right corner)
After powering up a few times, these resistors were blown. Admittedly, they were 10 Ohms instead of 1 Ohms. Nevertheless, I think we should not underestimate the peak power during startup, even in case of very small resistors, like 1 Ohm.
Cheers,
E.
Last edited:
Thick film is usually considerably more robust against inrush current situations (www.irctt.com/pdf/Resistors_Surge_Conditions_PMDL_Feb07_article.pdf). Also there are special parts rated for surge current limiting.
Samuel
Samuel