Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

[SOLVED] Design precaution of 78xx and 79xx series regulators

Status
Not open for further replies.
Thanks guys for the help. The reason I posted this to ask was because my experience on regulators and converters are mostly conditional stable only. They mostly have closed loop feedback to stabilize the output voltage. To every closed loop feedback system, the more the caps you put on the output, the more chance it can be unstable. You cannot even trust the application circuits. We had a body camera that kept giving slanted stripes on the screen, I found the two converter circuits that was designed in oscillated!!! The circuits were exactly given by Linear Technology application notes. They did NOT work. People have a big miss conception that power supply circuit are slow, that you don't have to treat it as a closed loop feedback system, to worry about the poles and zeros. That's the reason most of the SMPS has minimal load require and it will oscillate if you don't draw enough power.

That's the reason I want to confirm if it is ok to put extra caps as the datasheet only gives minimum value. I have not use this kind of regulator for like 30 years, I remember long time ago, I used a 7805, it got really hot even with no load, so I just not using them at all. I thought something wrong with the IC, but thinking back, it must be oscillation that generated so much heat. I have no recollection how I built the circuit, I was very green at the time.

Since everyone said it's ok here, I consider this is solved.

Thanks
 
Last edited:

A 7805 with no load shouldn't oscillate and shouldn't get hot unless there was something else wrong. Excessive input voltage, maybe?
 

A 7805 with no load shouldn't oscillate and shouldn't get hot unless there was something else wrong. Excessive input voltage, maybe?

It was a long time ago around 1979. I really don't remember what I did anymore. I didn't not have problem with the 78Lxx and 79Lxx, so I have been using those and never touch the regular ones. As I said, since all the experience of instability on SMPS and DC DC converters, I just want to be cautious and ask here.

Thanks
 

I worked in audio all my career. Sound "opens up" for people and animals who say they can hear the 20kHz sounds better when the 40kHz and 60kHz harmonics (that are not heard) are reduced.
Fundamental sounds must be very loud for 1% distortion (-40dB) to be heard but then the fundamental sound covers up the distortion harmonics.

Instead of saying "the sound opens up" I say "the sound is smooth". Instead of saying "the sound is closed" I say "the sound is rough".
 

I worked in audio all my career. Sound "opens up" for people and animals who say they can hear the 20kHz sounds better when the 40kHz and 60kHz harmonics (that are not heard) are reduced.
Fundamental sounds must be very loud for 1% distortion (-40dB) to be heard but then the fundamental sound covers up the distortion harmonics.

Instead of saying "the sound opens up" I say "the sound is smooth". Instead of saying "the sound is closed" I say "the sound is rough".

This is so subjective. Also, it's not easy to find people that are into audiophile. Most people just happy with some Yamaha, Pioneer low or mid line stuffs. And even harder to find people that know electronics and into audiophile.

I can only speak for my own experience. I am an electrical engineer for 30 years and manager of EE for over 10 years. But my career is not in audio, more an variety of analog, IC design, data acquisition, RF etc. I started my career designing guitar amps and started again after I retired. Then I move to designing audiophile power amps in the last 3 years. I gave up guitar amps after designed and built two of the channel switching, high gain tube amps because I don't perform anymore, it's no fun designing guitar amp that I don't use after it's completed.

I don't go out and audition equipments, for one, most of the high end audiophile stores are not around anymore, not like 10 or 20 year ago that I can go to different stores and listen. So I have limited speakers and amps that I can listen to. I have/had two pairs of Kef, one Uni-Q floor standing with dual woofers, one big book shelf. A pair of Monitor Audio book shelf and a pair of JM Lab Spectral 913.1 that is mid/upper line of JM Lab Focal. I have a Nakamichi PA-7 Stasis power amp that is almost identical to the Threshold S300 but slightly better ( with one more pairs of output transistors and newer 30MHz power transistors), all designed by Nelson Pass. I also have an Acurus 3X200. That's it!!! The rest are those cheaper Marantz and old Japanese SAE. So I can only make my comments within these few components.

You should know it's very hard to make blanket statement as the sound depends on the matching of speakers, cables and amplifiers. Some speakers are just not very picky, you can get away with just monster 12 gauge and lesser amps. Some speakers just very taxing on the speaker cables and amps. I can only say with my JM Lab, it is very critical to have very good speaker cable( just as important as good amp) and very low THD and high damping factor amps.

If you look into the schematic of all the power amps, their basic concept are very similar. Mostly are either follow the Blameless or complementary IPS design. Most use BJT. Those are all covered in the book by Bob Cordell. BUT most amp take short cuts, removing some of the features. Even the famous Krell KSA series are very primitive and not implementing a lot of the design in the book. They just have a lot of pairs of output transistors running in Class A to make up the short coming. The only reason I can think of is the more you put in, the harder it is to tame and get reliable stable circuit for production. Things like current source load for IPS LTP, darlington VAS, particular 3EF OPS. Last but not the least, many output transistor pairs. My first design was blameless with all the above with 9 pairs of output transistors. Everyone of these make it harder to tame, LTSpice is NOT accurate in simulation at all because of the large signal model is not accurate. Took me like two weeks to get it very stable with capacitor load test. I achieved the THD spec published by Bob Cordell. From blind test with a few people, it kicked the butt of both the Nakamichi and Acurus. And the comment is better imaging, more 3D, you can hear different instruments more clearly and like they are in different position in the room. I found it corresponding to the THD.

Power amp is just a big discrete opamp with huge drive. Both Bob Cordell and me were analog bipolar IC design engineers. I used to talk to him on forum before. I really can relate to his book because I designed IC opamps and it's very similar in certain ways.

Too bad you are in Canada, I would love to meet up with people that design amps and listen to each others' setup, what they think is important and what they think makes the difference. Nobody has wide spectrum of speakers and amps to cover everything to have the broad picture. Sharing experience is the only way.

Cheers.

Alan


BTW, what audio equipment you design? There is nothing for me even though I live in the heart of the electronic industry.....the Bay Area. It there is a high end audio company in my area, I sure don't mind coming out of retirement and work for a little. Hell, it cost me very close to $1000 to build one amp ( just parts, pcb and the most expensive is the chassis). It would be nice to at least don't have to pay to build an amp!!! Here are a few internal pictures of one of the amp I built.

Finished amp modules 1.jpg

Complete amp.JPG

ADD IPS filter 2.JPG
 
Last edited:

Many expensive speakers are "pretty". The high cost is in the looks, not in the sound.
Some amplifiers have vacuum tubes on top but they are not parts in the circuit, they are there for looks.

I have been retired for many years and was shocked recently when a free hearing test revealed that I have normal-for-my-age (72) high frequency hearing loss. Like double the highs cut when a tone control turned all the way down and I could not hear most distortion anymore. My new hearing aids returned high frequencies to my hearing and produce no distortion and no noise. But some audio systems obviously produce distortion with my hearing aids turned on. The hearing aids have selectable modes that normal hearing cannot do like extra sensitivity, muting, noise reduction and directionality. Of course they compress sounds that are too loud and the compression is barely noticeable without any distortion.
 

Many expensive speakers are "pretty". The high cost is in the looks, not in the sound.
Some amplifiers have vacuum tubes on top but they are not parts in the circuit, they are there for looks.

I have been retired for many years and was shocked recently when a free hearing test revealed that I have normal-for-my-age (72) high frequency hearing loss. Like double the highs cut when a tone control turned all the way down and I could not hear most distortion anymore. My new hearing aids returned high frequencies to my hearing and produce no distortion and no noise. But some audio systems obviously produce distortion with my hearing aids turned on. The hearing aids have selectable modes that normal hearing cannot do like extra sensitivity, muting, noise reduction and directionality. Of course they compress sounds that are too loud and the compression is barely noticeable without any distortion.

I don't have a lot of experience with speakers. I owned two pairs of Kefs and even one Reference Series Kef center speaker. I think Kef is way over rated. When I up graded to the JM Lab, it was day and night difference. I am actually surprised the Monitor Audio little bookshelf is quite good. My stepson gave them to me, I use it as decoration for a few years. When I was bringing up my amp built, I ddn't want to take the risk hooking up to the JM Lab, so I use the Monitor Audio as "fuse". But it was a present surprise. Even at this mid line speakers, the treble and bass are all good. To me, it's the separation, sound stage and transparency that I am looking for. Kef's name is just bigger than the quality to me. Another brand I like is Sonus Faber. Surprisingly, I tested one pair of B&W and I did not like it.

Ha ha, I never saw any amp that put tubes just for decoration!!!

I really want to find an excuse to get a hearing test. I used to have ( stressing USED TO) very good ears. I had a hearing test 30 years ago, my ears were so sensitive that they kept lowering the volume and asked whether I could hear them. They finally turned off the mic( I could hear the hiss stopped) and I could still answer their question over the sound proofing glass in the next room. High frequency sound really irritates me. It still irritates me!!! I never complain about system not having enough highs, always complain not enough lower mid and bass!!! I really would like to have a test to see where my hearing is now.

That's the big problem in judging sound quality, people have different ears, what sounds good for one might not be good for others. In the preamp I designed, I had to put in a +1dB or so boost below 500Hz. Just slightly warming the sound up.

Electronics is the passion of my life. I started out as a rock and funk guitarist in the 70s. I never happy with my amps and I got really good ones. So I started modifying guitar amps. I started using variac ONLY on the output and differential phase splitter and use separate filament and preamp supply to keep the voltages optimal. I had great success. But at the same time, I fell in love with electronics. I finally quit playing in 1979 all together and pursue electronics and never look back. I got into difference facet of electronics by switching jobs, from firm ware uP design, to data acquisition system with LeCroy, then analog IC design, then frontend of Ultrasound medical scanner with Siemens before settling in designing mass spectrometers for semiconductor analytical equipment for like 12 years, then I got restless and moved to Telecom and then to military stuffs before I retired in 2006. After I retire, I still keep going, now is audio power amps!!! When I get tired of audio amps, I'll move onto some other electronics. Hey, it's better than playing crossword puzzles!!!
 

I'm well into my 60's now but just before retiring people in the factory used to think I was mad (I probably was/am but by different criteria) because I cringed when I head a very loud high pitched whistle that they were oblivious to. Rather than doubt my sanity, I went to track it down and my hearing led me to an ultrasonic welder running at 20KHz. As the welder fixed windows into an enclosure, to me the noise was ear shattering. Everybody else, including much younger people heard nothing. Sadly, and I hope not because of the welder, I now suffer mild tinitus and always hear a crescendo of high pitched whistles so telling what is real and what's outside is a bit more difficult although I would still assess my hearing as pretty good.

I once argued with an audiophile that they weren't actually listening to the original sound waves anyway but the rate of change of sound pressure waves. They said I was nuts but couldn't find a way to explain how the voltage recovered from a tape head or magnetic record pick-up is proportional to the rate of change of magnetic flux or speed of stylus displacement rather than absolute amount. I won the argument. Please don't go into the rights or wrongs of my case as it leads to all kinds of problems with frequency and phase response and inevitably resolves to a discussion of how a capacitor or transformer manages to couple a signal anyway.

I admire your patience and skills a designing and building that amplifier. I've designed equipment for recording studios and TV stations but you show a devotion beyond normal call of duty.

Brian.
 
I'm well into my 60's now but just before retiring people in the factory used to think I was mad (I probably was/am but by different criteria) because I cringed when I head a very loud high pitched whistle that they were oblivious to. Rather than doubt my sanity, I went to track it down and my hearing led me to an ultrasonic welder running at 20KHz. As the welder fixed windows into an enclosure, to me the noise was ear shattering. Everybody else, including much younger people heard nothing. Sadly, and I hope not because of the welder, I now suffer mild tinitus and always hear a crescendo of high pitched whistles so telling what is real and what's outside is a bit more difficult although I would still assess my hearing as pretty good.

I once argued with an audiophile that they weren't actually listening to the original sound waves anyway but the rate of change of sound pressure waves. They said I was nuts but couldn't find a way to explain how the voltage recovered from a tape head or magnetic record pick-up is proportional to the rate of change of magnetic flux or speed of stylus displacement rather than absolute amount. I won the argument. Please don't go into the rights or wrongs of my case as it leads to all kinds of problems with frequency and phase response and inevitably resolves to a discussion of how a capacitor or transformer manages to couple a signal anyway.

I admire your patience and skills a designing and building that amplifier. I've designed equipment for recording studios and TV stations but you show a devotion beyond normal call of duty.

Brian.

Thanks for the complement. I want to do it right when I build the amps, no short cuts. All wires are either on terminal blocks or Molex connectors. I can take the whole amp apart with just one or two solder joints, other than that, just screw driver.

You bring up an interesting point about the rate of change. I always wonder and I even question whether our body can feel the high or low frequencies even though our ears cannot hear it. The sound pressure that make the difference on how we perceive as sound quality. To me it's the slew rate of the amp. I have no idea how valid is this, I am just wondering.

I design my amps with slew rate of about 30V/uS, and with small signal BW of over 350KHz. I've seen amps that spec BW of 1MHz and even faster slew rate. One thing I notice a good amp gives more impact on the sound, that you feel the sound even though the sound is not harsh. This applies to bass also even though slew rate is not an issue for low frequencies.

One thing I find that no manufacturers talk about, THD and low frequencies. I found THD goes up below 40Hz just like going to higher frequencies. Like at 20Hz, THD is just as if not higher than in 20KHz in my experience in testing a few amps. People that use closed loop feedback to zero out DC offset at the output ( more common than people think) better worry about adding distortion at low frequency when the loop kicks in at low frequency. I have one amp that I put the DC correction in and I had to really put the poll at very low frequency to take the feedback out of action at 20Hz.

I use sub woofer, when I change out my old Velodyne ( 20 years old, first generation type) to a Rhymic closed loop feed back type, the impact is so obvious that it even enhance the mid frequency and make door closing and things drop in tv shows more real. That I did not expect. All by changing to a newer technology sub. I was actually surprised with the improvement.

Of cause, these are just my impression, no scientific proof.

Alan
 

Hi,

I cringed when I head a very loud high pitched whistle that they were oblivious to. Rather than doubt my sanity, I went to track it down and my hearing led me to an ultrasonic welder running at 20KHz.

I'm so glad you've said that. For the second time living in one specific place - nowhere else has this happened (lived here happened, back to Blighty a few months and it stopped completely, back here and started again), I doubt that is my imagination somehow - every few months or so I feel/hear a high-pitched noise in one ear that grows and sort of hurts even, like an aural drill, it lasts from a few seconds to maybe 30 seconds or even a minute tops, I can't tell because it's horrible and feels like a directed weapon. I'd assumed that as I'm not 100% insane yet that presumably it must come from some machine operated nearby or one of the chicken farms or something. I can only guess it's something like the experience and machine you describe.

It's not the bells (of Notre-Dame) just yet for me then, it's a machine, phew...

Seriously, thanks, great explanation.
 

I want to find the voltage of the secondary of the transformer I need to buy for 7812 and 7805. I think I made a mistake of buying 6.3V secondary for 7805 and 15V secondary for 7812. I was thinking that will rectify out to 6.3 X 1.414 = 8.9Vpeak and 15 X 1.414 = 21.2Vpeak. But I think at least the 6.3V is too low for 7805.

I have 1000uF right after the bridge rectifier. But I think I am cutting close on the 12V and messed up on the 5V. I cannot afford to over kill and use a 6VA transformer because of the height restriction. 6VA is the only one I can get. I can get 8V secondary, that's about it. Might have to increase the cap after the bridge rectifier to keep the voltage from drooping too low.

Please comment.

Thanks
 
Last edited:

Texas Instruments recommends a minimum input of 7V for a 7805. The voltage drop of a silicon bridge rectifier is 2V and who knows how low your electricity drops when everyone in your neighbourhood turns on the stove or air conditioner.
So the peak should be at least 9V which is 6.3V RMS AC.
 
I try to do some calculation. From datasheet, drop out voltage is 2V, lets include 1V ripple, so I need minimum of 3V above the output voltage.

1) For 7812 ( also 7912), I need 15V at the input. For 30V center tapped ( 15V each side) running bridge rectifier. I only have to worry about one diode drop. 15VAC = 15 X 1.414 = 21.21. Let's call it 20V with load. Minus 1diode will be 19V. So I have 19-12 = 7V headroom. My conclusion is 15V-15V transformer is good for 7812 and 7912.

2) For 7805, 6.3V transformer gives 6.3 X 1.414 =8.9Vpeak. 2 diode drop ( 6.3V in parallel) take it down to 8.9-1.5 = 7.4V. With ripple of 1V, it's down to 6.4V. That's not good.

3) Using 8V transformer for 7805 gives 8 X 1.414 = 11.3V, 2 diode drop take it down to 9.8V. Lets say it's 9.3V loaded. 1V ripple drops it down to 8.3V. so I have 3.3V headroom for 7805. This should work out well.


Please tell me what you think.

Thanks

Alan
 

Hi,

You may use zener diodes to reduce voltage drop.

Additionally you should do some measurements on the real circuit, I assume you will see higher voltage than expected.
(No guarantee).

For my taste 7V of headroom for the 12V supply is a waste of power --> it just generates heat.
I'd choose a lower voltage transformer.
(Personal opinion)

Klaus
 
Hi,

You may use zener diodes to reduce voltage drop.

Additionally you should do some measurements on the real circuit, I assume you will see higher voltage than expected.
(No guarantee).

For my taste 7V of headroom for the 12V supply is a waste of power --> it just generates heat.
I'd choose a lower voltage transformer.
(Personal opinion)

Klaus

Thanks for the reply, so let's try 12V transformer

12V gives 12 X 1.414 = 17V, 1 diode drop becomes 16V, let's call it 15.5V under load. 1V ripple gets down to 14.5V. That gives 2.5V of headroom for 7812. That should work. BUT do you think this cut a little close?

Too bad they don't have 13V or 14V transformer.


PCB is out for fab and won't be back for another 10 days. So this is all empty talk. But I want to buy the transformers to cover all case right now.

Thanks

Alan
 

Hi,

Transformer voltage is for full load, it will be much higher on partial load.
Diode drop is for full load

But I assess you not to draw full load?

Klaus
 

... but as a general 'rule of thumb' figure, your calculations are about right. You need the bottom of the ripple on the reservoir capacitor to be just higher than the minimum input voltage of the regulator but taking into account the warning from Audioguru, you should allow maybe 1V or more as a safety margin. The transformer voltage and rectifier drop will result in a higher than expected voltage under low load current but that probably isn't an issue as the power dissipated is (Vin -Vout)/I soto some degree they offset each other.

Brian.
 
Thanks guys.

After thinking more, I think I am going to get the 8V transformer for the 7805 and keep the 15V transformer for the 7812. the circuit is low power, I don't want to deal with any slight chance of increasing noise ( low overhead). This is small amount heat compare to about 120W of heat generated by the amp itself. So far, the quality really requires a lot of heat, I don't know of any other way other than error correction at the output stage. This will be my next design, to do error correction on the crossover distortion so I can lower the idle current of the amp. I just don't think Class D or Class H is the way to go for high end audiophile design.
 

120W of heat!? Then you are making a class-A home heater, not a modern audio amplifier that has a small class-AB idle current and lots of negative feedback for extremely low distortion. The negative feedback and wide bandwidth does the error correction like most modern hifi amps.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top