tantalum
I test tantalums by production date. More than 10 years old, suspect. The capacitors that started as tantalums in my 1968 production Hammond organ have all random dates, starting in 1973, so some of them were bad after 5 years. There were some 5 & 10 uf tantalums in there originally, I replaced them all with Aerovox ceramic caps. The +80 -20 tolerance on the ceramics had no bad side effects I can detect . There is no reason besides first cost to use an electrolytic or tantalum capacitor in 1`,2,3,5,10 uf below 50 WVDC unless it is a critical timing or filtration circuit.
I test tantalums by production date. More than 10 years old, suspect. The capacitors that started as tantalums in my 1968 production Hammond organ have all random dates, starting in 1973, so some of them were bad after 5 years. There were some 5 & 10 uf tantalums in there originally, I replaced them all with Aerovox ceramic caps. The +80 -20 tolerance on the ceramics had no bad side effects I can detect . There is no reason besides first cost to use an electrolytic or tantalum capacitor in 1`,2,3,5,10 uf below 50 WVDC unless it is a critical timing or filtration circuit.
Last edited:
capacitor purchase
If you matched ESR, size, tolerance, voltage rating, I don't know why I would ever buy another tantalum capacitor. They were a lot smaller than aluminum electrolytics in the seventies, but aluminum is 1/3 the size now. Since I keep my favorite equipment for decades (repairability is one criterea that most modern stuff fails, I can't wave solder) and I don't like doing work over again in 10 years, I buy aluminum electrolytics by projected service life. 5000 hours up at 105 deg is good, 500 hours or less at 80 deg bad. I notice when I get the long life ones they have nice looking seals around the leads. Really cheap electrolytics are not rated in hours, nor are American made Vishay Sprague, (TL) which I am partial to since some built by the predecessor company are still working fine in the harp circuit of my organ after 40 years. One vendor that shows the service life in their selector tables is Newark/Farnell. They also have a discount mark next to the cap if it is overstock or old- old is bad on electrolytics. Since the size of the cap can be gotten from the manufacturer's link in the on-line catalog, you can tell if an aluminum electrolytic will fit on your board before you order it. Watch tolerance on timing circuits, critical point frequency filters. For power supply filtering or bypassing hum to return at an input of an amp stage, +80-20 seems to work fine. The harp capacitors were 15 uf +_10% is one reason I didn't replace them, $5 each and value is important. I just bought a sack of 47 @ 25 V long life vishay spragues caps for a mixer that came in US made, to my delight as an unemployed former electronics specialist. Pro-grade motor drives mostly use Nichon electrolytic caps, 10 years next to an oven at high currents proves these expensive caps were really built.
Oh, self inductance measurement is important in caps for RF circuits, not very important for audio. Watch this if you are doing RF or fast computer circuits, one reason for tantalums that doesn't apply to the organs and sound equpment I am doing.
If you matched ESR, size, tolerance, voltage rating, I don't know why I would ever buy another tantalum capacitor. They were a lot smaller than aluminum electrolytics in the seventies, but aluminum is 1/3 the size now. Since I keep my favorite equipment for decades (repairability is one criterea that most modern stuff fails, I can't wave solder) and I don't like doing work over again in 10 years, I buy aluminum electrolytics by projected service life. 5000 hours up at 105 deg is good, 500 hours or less at 80 deg bad. I notice when I get the long life ones they have nice looking seals around the leads. Really cheap electrolytics are not rated in hours, nor are American made Vishay Sprague, (TL) which I am partial to since some built by the predecessor company are still working fine in the harp circuit of my organ after 40 years. One vendor that shows the service life in their selector tables is Newark/Farnell. They also have a discount mark next to the cap if it is overstock or old- old is bad on electrolytics. Since the size of the cap can be gotten from the manufacturer's link in the on-line catalog, you can tell if an aluminum electrolytic will fit on your board before you order it. Watch tolerance on timing circuits, critical point frequency filters. For power supply filtering or bypassing hum to return at an input of an amp stage, +80-20 seems to work fine. The harp capacitors were 15 uf +_10% is one reason I didn't replace them, $5 each and value is important. I just bought a sack of 47 @ 25 V long life vishay spragues caps for a mixer that came in US made, to my delight as an unemployed former electronics specialist. Pro-grade motor drives mostly use Nichon electrolytic caps, 10 years next to an oven at high currents proves these expensive caps were really built.
Oh, self inductance measurement is important in caps for RF circuits, not very important for audio. Watch this if you are doing RF or fast computer circuits, one reason for tantalums that doesn't apply to the organs and sound equpment I am doing.
Last edited:
I looked at spec sheet on the regulator circuit & it's recomended a 1uf tantalum or a 10uf electrolytic(not sure why the large difference). The caps are used for stability of the regulator. the 47uf is in a timing ciruit in other part of machine. A 10uf electrolytic is still small so guess I'm in business.
You bring up a good point .
There seam to have been two reasons to use tant's
1. low leakage for timing uses.
there are other alternatives now.
2. small size.
Once again other alternatives now.
Bottom line better of due to their reputation and the high level of toxicity of tantalum, replace them all, (my advice )
There seam to have been two reasons to use tant's
1. low leakage for timing uses.
there are other alternatives now.
2. small size.
Once again other alternatives now.
Bottom line better of due to their reputation and the high level of toxicity of tantalum, replace them all, (my advice )
There's been a belief that tantalums have some magic low impedance property that's good for stabilizing regulators. Not true, but the esr does tend to fall in a narrow range, so the performance is very predictable. More so than aluminum electrolytics, and much higher than ceramic. Other than that, size is their only advantage- an advantage lost to modern aluminum electrolytics. I don't find much trouble with quality tantalums but would never design one into a new circuit. I don't have any problem replacing a bad one with another tantalum, and would examine the usage carefully before using a lower impedance aluminum electrolytic or ceramic. Remember they can't tolerate negative voltages. Installed backwards they usually short out and get hot. More annoying, if they don't short out, they become a noise generator. When I see a regulator with a noisy output, the first thing I look for is a defective or backwards tantalum.
Actually there is a very good reason why tants are used, especially in regulator circuits - and that's because they have high ESRs. Usually in the 1-10 ohm range this extra resistance damps any resonance with the slightly inductive output impedance of the regulator. This is also the reason why very large values of capacitor are not usually used on the output of regulators. This resonance then moves into the audio band and reduces the available regulation rather than improving it as one might expect.
Testing Tantalum Caps
If you want to test Ta caps (or any other passive component for that matter) the best tool is a vector network analyzer (or VNA). this tool can characterize the real and imaginary components of insertion or reflection loss as a function of frequency. It will easily identify opens and shorts and can characterize frequency dependent effects such as self resonance. VNAs come in different frequency ranges, and for audio work you want one that goes down to ~50 Hz and up to ~50 MHz. One nice attribute of VNAs is the ability to calibrate out the parasitics of the cabling and connectors.
If you want to test Ta caps (or any other passive component for that matter) the best tool is a vector network analyzer (or VNA). this tool can characterize the real and imaginary components of insertion or reflection loss as a function of frequency. It will easily identify opens and shorts and can characterize frequency dependent effects such as self resonance. VNAs come in different frequency ranges, and for audio work you want one that goes down to ~50 Hz and up to ~50 MHz. One nice attribute of VNAs is the ability to calibrate out the parasitics of the cabling and connectors.
VNA Cost
I have seen several Agilent HP 3577A Network Analyzers with a 5Hz-200MHz BW for sale on Ebay in $1000 range. Newer VNAs cost from $20K - $150K, depending on the options, but these typically have bandwidths of 25 GHz or greater, not much use for audio unless you need to measure the IC package insertion loss at 20 GHz.
I have seen several Agilent HP 3577A Network Analyzers with a 5Hz-200MHz BW for sale on Ebay in $1000 range. Newer VNAs cost from $20K - $150K, depending on the options, but these typically have bandwidths of 25 GHz or greater, not much use for audio unless you need to measure the IC package insertion loss at 20 GHz.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.